TY - CONF AB - Deep neural networks (DNNs) often have to be compressed, via pruning and/or quantization, before they can be deployed in practical settings. In this work we propose a new compression-aware minimizer dubbed CrAM that modifies the optimization step in a principled way, in order to produce models whose local loss behavior is stable under compression operations such as pruning. Thus, dense models trained via CrAM should be compressible post-training, in a single step, without significant accuracy loss. Experimental results on standard benchmarks, such as residual networks for ImageNet classification and BERT models for language modelling, show that CrAM produces dense models that can be more accurate than the standard SGD/Adam-based baselines, but which are stable under weight pruning: specifically, we can prune models in one-shot to 70-80% sparsity with almost no accuracy loss, and to 90% with reasonable (∼1%) accuracy loss, which is competitive with gradual compression methods. Additionally, CrAM can produce sparse models which perform well for transfer learning, and it also works for semi-structured 2:4 pruning patterns supported by GPU hardware. The code for reproducing the results is available at this https URL . AU - Peste, Elena-Alexandra AU - Vladu, Adrian AU - Kurtic, Eldar AU - Lampert, Christoph AU - Alistarh, Dan-Adrian ID - 13053 T2 - 11th International Conference on Learning Representations TI - CrAM: A Compression-Aware Minimizer ER - TY - THES AB - Deep learning has become an integral part of a large number of important applications, and many of the recent breakthroughs have been enabled by the ability to train very large models, capable to capture complex patterns and relationships from the data. At the same time, the massive sizes of modern deep learning models have made their deployment to smaller devices more challenging; this is particularly important, as in many applications the users rely on accurate deep learning predictions, but they only have access to devices with limited memory and compute power. One solution to this problem is to prune neural networks, by setting as many of their parameters as possible to zero, to obtain accurate sparse models with lower memory footprint. Despite the great research progress in obtaining sparse models that preserve accuracy, while satisfying memory and computational constraints, there are still many challenges associated with efficiently training sparse models, as well as understanding their generalization properties. The focus of this thesis is to investigate how the training process of sparse models can be made more efficient, and to understand the differences between sparse and dense models in terms of how well they can generalize to changes in the data distribution. We first study a method for co-training sparse and dense models, at a lower cost compared to regular training. With our method we can obtain very accurate sparse networks, and dense models that can recover the baseline accuracy. Furthermore, we are able to more easily analyze the differences, at prediction level, between the sparse-dense model pairs. Next, we investigate the generalization properties of sparse neural networks in more detail, by studying how well different sparse models trained on a larger task can adapt to smaller, more specialized tasks, in a transfer learning scenario. Our analysis across multiple pruning methods and sparsity levels reveals that sparse models provide features that can transfer similarly to or better than the dense baseline. However, the choice of the pruning method plays an important role, and can influence the results when the features are fixed (linear finetuning), or when they are allowed to adapt to the new task (full finetuning). Using sparse models with fixed masks for finetuning on new tasks has an important practical advantage, as it enables training neural networks on smaller devices. However, one drawback of current pruning methods is that the entire training cycle has to be repeated to obtain the initial sparse model, for every sparsity target; in consequence, the entire training process is costly and also multiple models need to be stored. In the last part of the thesis we propose a method that can train accurate dense models that are compressible in a single step, to multiple sparsity levels, without additional finetuning. Our method results in sparse models that can be competitive with existing pruning methods, and which can also successfully generalize to new tasks. AU - Peste, Elena-Alexandra ID - 13074 SN - 2663-337X TI - Efficiency and generalization of sparse neural networks ER - TY - JOUR AB - The development of two-dimensional materials has resulted in a diverse range of novel, high-quality compounds with increasing complexity. A key requirement for a comprehensive quantitative theory is the accurate determination of these materials' band structure parameters. However, this task is challenging due to the intricate band structures and the indirect nature of experimental probes. In this work, we introduce a general framework to derive band structure parameters from experimental data using deep neural networks. We applied our method to the penetration field capacitance measurement of trilayer graphene, an effective probe of its density of states. First, we demonstrate that a trained deep network gives accurate predictions for the penetration field capacitance as a function of tight-binding parameters. Next, we use the fast and accurate predictions from the trained network to automatically determine tight-binding parameters directly from experimental data, with extracted parameters being in a good agreement with values in the literature. We conclude by discussing potential applications of our method to other materials and experimental techniques beyond penetration field capacitance. AU - Henderson, Paul M AU - Ghazaryan, Areg AU - Zibrov, Alexander A. AU - Young, Andrea F. AU - Serbyn, Maksym ID - 14320 IS - 12 JF - Physical Review B SN - 2469-9950 TI - Deep learning extraction of band structure parameters from density of states: A case study on trilayer graphene VL - 108 ER - TY - CONF AB - This paper focuses on the implementation details of the baseline methods and a recent lightweight conditional model extrapolation algorithm LIMES [5] for streaming data under class-prior shift. LIMES achieves superior performance over the baseline methods, especially concerning the minimum-across-day accuracy, which is important for the users of the system. In this work, the key measures to facilitate reproducibility and enhance the credibility of the results are described. AU - Tomaszewska, Paulina AU - Lampert, Christoph ID - 14410 SN - 0302-9743 T2 - International Workshop on Reproducible Research in Pattern Recognition TI - On the implementation of baselines and lightweight conditional model extrapolation (LIMES) under class-prior shift VL - 14068 ER - TY - JOUR AB - Recent work has paid close attention to the first principle of Granger causality, according to which cause precedes effect. In this context, the question may arise whether the detected direction of causality also reverses after the time reversal of unidirectionally coupled data. Recently, it has been shown that for unidirectionally causally connected autoregressive (AR) processes X → Y, after time reversal of data, the opposite causal direction Y → X is indeed detected, although typically as part of the bidirectional X↔ Y link. As we argue here, the answer is different when the measured data are not from AR processes but from linked deterministic systems. When the goal is the usual forward data analysis, cross-mapping-like approaches correctly detect X → Y, while Granger causality-like approaches, which should not be used for deterministic time series, detect causal independence X → Y. The results of backward causal analysis depend on the predictability of the reversed data. Unlike AR processes, observables from deterministic dynamical systems, even complex nonlinear ones, can be predicted well forward, while backward predictions can be difficult (notably when the time reversal of a function leads to one-to-many relations). To address this problem, we propose an approach based on models that provide multiple candidate predictions for the target, combined with a loss function that consideres only the best candidate. The resulting good forward and backward predictability supports the view that unidirectionally causally linked deterministic dynamical systems X → Y can be expected to detect the same link both before and after time reversal. AU - Jakubík, Jozef AU - Bui Thi Mai, Phuong AU - Chvosteková, Martina AU - Krakovská, Anna ID - 14446 IS - 4 JF - Measurement Science Review TI - Against the flow of time with multi-output models VL - 23 ER - TY - CONF AB - Pruning—that is, setting a significant subset of the parameters of a neural network to zero—is one of the most popular methods of model compression. Yet, several recent works have raised the issue that pruning may induce or exacerbate bias in the output of the compressed model. Despite existing evidence for this phenomenon, the relationship between neural network pruning and induced bias is not well-understood. In this work, we systematically investigate and characterize this phenomenon in Convolutional Neural Networks for computer vision. First, we show that it is in fact possible to obtain highly-sparse models, e.g. with less than 10% remaining weights, which do not decrease in accuracy nor substantially increase in bias when compared to dense models. At the same time, we also find that, at higher sparsities, pruned models exhibit higher uncertainty in their outputs, as well as increased correlations, which we directly link to increased bias. We propose easy-to-use criteria which, based only on the uncompressed model, establish whether bias will increase with pruning, and identify the samples most susceptible to biased predictions post-compression. Our code can be found at https://github.com/IST-DASLab/pruned-vision-model-bias. AU - Iofinova, Eugenia B AU - Peste, Elena-Alexandra AU - Alistarh, Dan-Adrian ID - 14771 T2 - 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition TI - Bias in pruned vision models: In-depth analysis and countermeasures ER - TY - CONF AB - Neural collapse (NC) refers to the surprising structure of the last layer of deep neural networks in the terminal phase of gradient descent training. Recently, an increasing amount of experimental evidence has pointed to the propagation of NC to earlier layers of neural networks. However, while the NC in the last layer is well studied theoretically, much less is known about its multi-layered counterpart - deep neural collapse (DNC). In particular, existing work focuses either on linear layers or only on the last two layers at the price of an extra assumption. Our paper fills this gap by generalizing the established analytical framework for NC - the unconstrained features model - to multiple non-linear layers. Our key technical contribution is to show that, in a deep unconstrained features model, the unique global optimum for binary classification exhibits all the properties typical of DNC. This explains the existing experimental evidence of DNC. We also empirically show that (i) by optimizing deep unconstrained features models via gradient descent, the resulting solution agrees well with our theory, and (ii) trained networks recover the unconstrained features suitable for the occurrence of DNC, thus supporting the validity of this modeling principle. AU - Súkeník, Peter AU - Mondelli, Marco AU - Lampert, Christoph ID - 14921 T2 - 37th Annual Conference on Neural Information Processing Systems TI - Deep neural collapse is provably optimal for the deep unconstrained features model ER - TY - GEN AB - A crucial property for achieving secure, trustworthy and interpretable deep learning systems is their robustness: small changes to a system's inputs should not result in large changes to its outputs. Mathematically, this means one strives for networks with a small Lipschitz constant. Several recent works have focused on how to construct such Lipschitz networks, typically by imposing constraints on the weight matrices. In this work, we study an orthogonal aspect, namely the role of the activation function. We show that commonly used activation functions, such as MaxMin, as well as all piece-wise linear ones with two segments unnecessarily restrict the class of representable functions, even in the simplest one-dimensional setting. We furthermore introduce the new N-activation function that is provably more expressive than currently popular activation functions. We provide code at this https URL. AU - Prach, Bernd AU - Lampert, Christoph ID - 15039 T2 - arXiv TI - 1-Lipschitz neural networks are more expressive with N-activations ER - TY - GEN AB - We present Cross-Client Label Propagation(XCLP), a new method for transductive federated learning. XCLP estimates a data graph jointly from the data of multiple clients and computes labels for the unlabeled data by propagating label information across the graph. To avoid clients having to share their data with anyone, XCLP employs two cryptographically secure protocols: secure Hamming distance computation and secure summation. We demonstrate two distinct applications of XCLP within federated learning. In the first, we use it in a one-shot way to predict labels for unseen test points. In the second, we use it to repeatedly pseudo-label unlabeled training data in a federated semi-supervised setting. Experiments on both real federated and standard benchmark datasets show that in both applications XCLP achieves higher classification accuracy than alternative approaches. AU - Scott, Jonathan A AU - Yeo, Michelle X AU - Lampert, Christoph ID - 12660 T2 - arXiv TI - Cross-client Label Propagation for transductive federated learning ER - TY - GEN AB - Modern machine learning tasks often require considering not just one but multiple objectives. For example, besides the prediction quality, this could be the efficiency, robustness or fairness of the learned models, or any of their combinations. Multi-objective learning offers a natural framework for handling such problems without having to commit to early trade-offs. Surprisingly, statistical learning theory so far offers almost no insight into the generalization properties of multi-objective learning. In this work, we make first steps to fill this gap: we establish foundational generalization bounds for the multi-objective setting as well as generalization and excess bounds for learning with scalarizations. We also provide the first theoretical analysis of the relation between the Pareto-optimal sets of the true objectives and the Pareto-optimal sets of their empirical approximations from training data. In particular, we show a surprising asymmetry: all Pareto-optimal solutions can be approximated by empirically Pareto-optimal ones, but not vice versa. AU - Súkeník, Peter AU - Lampert, Christoph ID - 12662 T2 - arXiv TI - Generalization in Multi-objective machine learning ER - TY - JOUR AB - Fairness-aware learning aims at constructing classifiers that not only make accurate predictions, but also do not discriminate against specific groups. It is a fast-growing area of machine learning with far-reaching societal impact. However, existing fair learning methods are vulnerable to accidental or malicious artifacts in the training data, which can cause them to unknowingly produce unfair classifiers. In this work we address the problem of fair learning from unreliable training data in the robust multisource setting, where the available training data comes from multiple sources, a fraction of which might not be representative of the true data distribution. We introduce FLEA, a filtering-based algorithm that identifies and suppresses those data sources that would have a negative impact on fairness or accuracy if they were used for training. As such, FLEA is not a replacement of prior fairness-aware learning methods but rather an augmentation that makes any of them robust against unreliable training data. We show the effectiveness of our approach by a diverse range of experiments on multiple datasets. Additionally, we prove formally that –given enough data– FLEA protects the learner against corruptions as long as the fraction of affected data sources is less than half. Our source code and documentation are available at https://github.com/ISTAustria-CVML/FLEA. AU - Iofinova, Eugenia B AU - Konstantinov, Nikola H AU - Lampert, Christoph ID - 12495 JF - Transactions on Machine Learning Research SN - 2835-8856 TI - FLEA: Provably robust fair multisource learning from unreliable training data ER - TY - CONF AB - It is a highly desirable property for deep networks to be robust against small input changes. One popular way to achieve this property is by designing networks with a small Lipschitz constant. In this work, we propose a new technique for constructing such Lipschitz networks that has a number of desirable properties: it can be applied to any linear network layer (fully-connected or convolutional), it provides formal guarantees on the Lipschitz constant, it is easy to implement and efficient to run, and it can be combined with any training objective and optimization method. In fact, our technique is the first one in the literature that achieves all of these properties simultaneously. Our main contribution is a rescaling-based weight matrix parametrization that guarantees each network layer to have a Lipschitz constant of at most 1 and results in the learned weight matrices to be close to orthogonal. Hence we call such layers almost-orthogonal Lipschitz (AOL). Experiments and ablation studies in the context of image classification with certified robust accuracy confirm that AOL layers achieve results that are on par with most existing methods. Yet, they are simpler to implement and more broadly applicable, because they do not require computationally expensive matrix orthogonalization or inversion steps as part of the network architecture. We provide code at https://github.com/berndprach/AOL. AU - Prach, Bernd AU - Lampert, Christoph ID - 11839 SN - 9783031198021 T2 - Computer Vision – ECCV 2022 TI - Almost-orthogonal layers for efficient general-purpose Lipschitz networks VL - 13681 ER - TY - CONF AB - The digitalization of almost all aspects of our everyday lives has led to unprecedented amounts of data being freely available on the Internet. In particular social media platforms provide rich sources of user-generated data, though typically in unstructured form, and with high diversity, such as written in many different languages. Automatically identifying meaningful information in such big data resources and extracting it efficiently is one of the ongoing challenges of our time. A common step for this is sentiment analysis, which forms the foundation for tasks such as opinion mining or trend prediction. Unfortunately, publicly available tools for this task are almost exclusively available for English-language texts. Consequently, a large fraction of the Internet users, who do not communicate in English, are ignored in automatized studies, a phenomenon called rare-language discrimination.In this work we propose a technique to overcome this problem by a truly multi-lingual model, which can be trained automatically without linguistic knowledge or even the ability to read the many target languages. The main step is to combine self-annotation, specifically the use of emoticons as a proxy for labels, with multi-lingual sentence representations.To evaluate our method we curated several large datasets from data obtained via the free Twitter streaming API. The results show that our proposed multi-lingual training is able to achieve sentiment predictions at the same quality level for rare languages as for frequent ones, and in particular clearly better than what mono-lingual training achieves on the same data. AU - Lampert, Jasmin AU - Lampert, Christoph ID - 10752 SN - 9781665439022 T2 - 2021 IEEE International Conference on Big Data TI - Overcoming rare-language discrimination in multi-lingual sentiment analysis ER - TY - CONF AB - We introduce LIMES, a new method for learning with non-stationary streaming data, inspired by the recent success of meta-learning. The main idea is not to attempt to learn a single classifier that would have to work well across all occurring data distributions, nor many separate classifiers, but to exploit a hybrid strategy: we learn a single set of model parameters from which a specific classifier for any specific data distribution is derived via classifier adaptation. Assuming a multiclass classification setting with class-prior shift, the adaptation step can be performed analytically with only the classifier’s bias terms being affected. Another contribution of our work is an extrapolation step that predicts suitable adaptation parameters for future time steps based on the previous data. In combination, we obtain a lightweight procedure for learning from streaming data with varying class distribution that adds no trainable parameters and almost no memory or computational overhead compared to training a single model. Experiments on a set of exemplary tasks using Twitter data show that LIMES achieves higher accuracy than alternative approaches, especially with respect to the relevant real-world metric of lowest within-day accuracy. AU - Tomaszewska, Paulina AU - Lampert, Christoph ID - 12161 T2 - 26th International Conference on Pattern Recognition TI - Lightweight conditional model extrapolation for streaming data under class-prior shift VL - 2022 ER - TY - CONF AB - Transfer learning is a classic paradigm by which models pretrained on large “upstream” datasets are adapted to yield good results on “downstream” specialized datasets. Generally, more accurate models on the “upstream” dataset tend to provide better transfer accuracy “downstream”. In this work, we perform an in-depth investigation of this phenomenon in the context of convolutional neural networks (CNNs) trained on the ImageNet dataset, which have been pruned-that is, compressed by sparsifiying their connections. We consider transfer using unstructured pruned models obtained by applying several state-of-the-art pruning methods, including magnitude-based, second-order, regrowth, lottery-ticket, and regularization approaches, in the context of twelve standard transfer tasks. In a nutshell, our study shows that sparse models can match or even outperform the transfer performance of dense models, even at high sparsities, and, while doing so, can lead to significant inference and even training speedups. At the same time, we observe and analyze significant differences in the behaviour of different pruning methods. The code is available at: https://github.com/IST-DASLab/sparse-imagenet-transfer. AU - Iofinova, Eugenia B AU - Peste, Elena-Alexandra AU - Kurtz, Mark AU - Alistarh, Dan-Adrian ID - 12299 T2 - 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition TI - How well do sparse ImageNet models transfer? ER - TY - JOUR AB - Addressing fairness concerns about machine learning models is a crucial step towards their long-term adoption in real-world automated systems. While many approaches have been developed for training fair models from data, little is known about the robustness of these methods to data corruption. In this work we consider fairness-aware learning under worst-case data manipulations. We show that an adversary can in some situations force any learner to return an overly biased classifier, regardless of the sample size and with or without degrading accuracy, and that the strength of the excess bias increases for learning problems with underrepresented protected groups in the data. We also prove that our hardness results are tight up to constant factors. To this end, we study two natural learning algorithms that optimize for both accuracy and fairness and show that these algorithms enjoy guarantees that are order-optimal in terms of the corruption ratio and the protected groups frequencies in the large data limit. AU - Konstantinov, Nikola H AU - Lampert, Christoph ID - 10802 JF - Journal of Machine Learning Research KW - Fairness KW - robustness KW - data poisoning KW - trustworthy machine learning KW - PAC learning SN - 1532-4435 TI - Fairness-aware PAC learning from corrupted data VL - 23 ER - TY - CONF AB - Addressing fairness concerns about machine learning models is a crucial step towards their long-term adoption in real-world automated systems. Many approaches for training fair models from data have been developed and an implicit assumption about such algorithms is that they are able to recover a fair model, despite potential historical biases in the data. In this work we show a number of impossibility results that indicate that there is no learning algorithm that can recover a fair model when a proportion of the dataset is subject to arbitrary manipulations. Specifically, we prove that there are situations in which an adversary can force any learner to return a biased classifier, with or without degrading accuracy, and that the strength of this bias increases for learning problems with underrepresented protected groups in the data. Our results emphasize on the importance of studying further data corruption models of various strength and of establishing stricter data collection practices for fairness-aware learning. AU - Konstantinov, Nikola H AU - Lampert, Christoph ID - 13241 T2 - Proceedings of Machine Learning Research TI - On the impossibility of fairness-aware learning from corrupted data VL - 171 ER - TY - THES AB - Because of the increasing popularity of machine learning methods, it is becoming important to understand the impact of learned components on automated decision-making systems and to guarantee that their consequences are beneficial to society. In other words, it is necessary to ensure that machine learning is sufficiently trustworthy to be used in real-world applications. This thesis studies two properties of machine learning models that are highly desirable for the sake of reliability: robustness and fairness. In the first part of the thesis we study the robustness of learning algorithms to training data corruption. Previous work has shown that machine learning models are vulnerable to a range of training set issues, varying from label noise through systematic biases to worst-case data manipulations. This is an especially relevant problem from a present perspective, since modern machine learning methods are particularly data hungry and therefore practitioners often have to rely on data collected from various external sources, e.g. from the Internet, from app users or via crowdsourcing. Naturally, such sources vary greatly in the quality and reliability of the data they provide. With these considerations in mind, we study the problem of designing machine learning algorithms that are robust to corruptions in data coming from multiple sources. We show that, in contrast to the case of a single dataset with outliers, successful learning within this model is possible both theoretically and practically, even under worst-case data corruptions. The second part of this thesis deals with fairness-aware machine learning. There are multiple areas where machine learning models have shown promising results, but where careful considerations are required, in order to avoid discrimanative decisions taken by such learned components. Ensuring fairness can be particularly challenging, because real-world training datasets are expected to contain various forms of historical bias that may affect the learning process. In this thesis we show that data corruption can indeed render the problem of achieving fairness impossible, by tightly characterizing the theoretical limits of fair learning under worst-case data manipulations. However, assuming access to clean data, we also show how fairness-aware learning can be made practical in contexts beyond binary classification, in particular in the challenging learning to rank setting. AU - Konstantinov, Nikola H ID - 10799 KW - robustness KW - fairness KW - machine learning KW - PAC learning KW - adversarial learning SN - 2663-337X TI - Robustness and fairness in machine learning ER - TY - CONF AB - Modern neural networks can easily fit their training set perfectly. Surprisingly, despite being “overfit” in this way, they tend to generalize well to future data, thereby defying the classic bias–variance trade-off of machine learning theory. Of the many possible explanations, a prevalent one is that training by stochastic gradient descent (SGD) imposes an implicit bias that leads it to learn simple functions, and these simple functions generalize well. However, the specifics of this implicit bias are not well understood. In this work, we explore the smoothness conjecture which states that SGD is implicitly biased towards learning functions that are smooth. We propose several measures to formalize the intuitive notion of smoothness, and we conduct experiments to determine whether SGD indeed implicitly optimizes for these measures. Our findings rule out the possibility that smoothness measures based on first-order derivatives are being implicitly enforced. They are supportive, though, of the smoothness conjecture for measures based on second-order derivatives. AU - Volhejn, Vaclav AU - Lampert, Christoph ID - 9210 SN - 0302-9743 T2 - 42nd German Conference on Pattern Recognition TI - Does SGD implicitly optimize for smoothness? VL - 12544 ER - TY - CONF AB - We study the inductive bias of two-layer ReLU networks trained by gradient flow. We identify a class of easy-to-learn (`orthogonally separable') datasets, and characterise the solution that ReLU networks trained on such datasets converge to. Irrespective of network width, the solution turns out to be a combination of two max-margin classifiers: one corresponding to the positive data subset and one corresponding to the negative data subset. The proof is based on the recently introduced concept of extremal sectors, for which we prove a number of properties in the context of orthogonal separability. In particular, we prove stationarity of activation patterns from some time onwards, which enables a reduction of the ReLU network to an ensemble of linear subnetworks. AU - Bui Thi Mai, Phuong AU - Lampert, Christoph ID - 9416 T2 - 9th International Conference on Learning Representations TI - The inductive bias of ReLU networks on orthogonally separable data ER - TY - GEN AB - Given the abundance of applications of ranking in recent years, addressing fairness concerns around automated ranking systems becomes necessary for increasing the trust among end-users. Previous work on fair ranking has mostly focused on application-specific fairness notions, often tailored to online advertising, and it rarely considers learning as part of the process. In this work, we show how to transfer numerous fairness notions from binary classification to a learning to rank setting. Our formalism allows us to design methods for incorporating fairness objectives with provable generalization guarantees. An extensive experimental evaluation shows that our method can improve ranking fairness substantially with no or only little loss of model quality. AU - Konstantinov, Nikola H AU - Lampert, Christoph ID - 10803 T2 - arXiv TI - Fairness through regularization for learning to rank ER - TY - THES AB - Deep learning is best known for its empirical success across a wide range of applications spanning computer vision, natural language processing and speech. Of equal significance, though perhaps less known, are its ramifications for learning theory: deep networks have been observed to perform surprisingly well in the high-capacity regime, aka the overfitting or underspecified regime. Classically, this regime on the far right of the bias-variance curve is associated with poor generalisation; however, recent experiments with deep networks challenge this view. This thesis is devoted to investigating various aspects of underspecification in deep learning. First, we argue that deep learning models are underspecified on two levels: a) any given training dataset can be fit by many different functions, and b) any given function can be expressed by many different parameter configurations. We refer to the second kind of underspecification as parameterisation redundancy and we precisely characterise its extent. Second, we characterise the implicit criteria (the inductive bias) that guide learning in the underspecified regime. Specifically, we consider a nonlinear but tractable classification setting, and show that given the choice, neural networks learn classifiers with a large margin. Third, we consider learning scenarios where the inductive bias is not by itself sufficient to deal with underspecification. We then study different ways of ‘tightening the specification’: i) In the setting of representation learning with variational autoencoders, we propose a hand- crafted regulariser based on mutual information. ii) In the setting of binary classification, we consider soft-label (real-valued) supervision. We derive a generalisation bound for linear networks supervised in this way and verify that soft labels facilitate fast learning. Finally, we explore an application of soft-label supervision to the training of multi-exit models. AU - Bui Thi Mai, Phuong ID - 9418 SN - 2663-337X TI - Underspecification in deep learning ER - TY - CHAP AB - The goal of zero-shot learning is to construct a classifier that can identify object classes for which no training examples are available. When training data for some of the object classes is available but not for others, the name generalized zero-shot learning is commonly used. In a wider sense, the phrase zero-shot is also used to describe other machine learning-based approaches that require no training data from the problem of interest, such as zero-shot action recognition or zero-shot machine translation. AU - Lampert, Christoph ED - Ikeuchi, Katsushi ID - 14987 SN - 9783030634155 T2 - Computer Vision TI - Zero-Shot Learning ER - TY - GEN AB - We present a generative model of images that explicitly reasons over the set of objects they show. Our model learns a structured latent representation that separates objects from each other and from the background; unlike prior works, it explicitly represents the 2D position and depth of each object, as well as an embedding of its segmentation mask and appearance. The model can be trained from images alone in a purely unsupervised fashion without the need for object masks or depth information. Moreover, it always generates complete objects, even though a significant fraction of training images contain occlusions. Finally, we show that our model can infer decompositions of novel images into their constituent objects, including accurate prediction of depth ordering and segmentation of occluded parts. AU - Anciukevicius, Titas AU - Lampert, Christoph AU - Henderson, Paul M ID - 8063 T2 - arXiv TI - Object-centric image generation with factored depths, locations, and appearances ER - TY - CONF AB - A natural approach to generative modeling of videos is to represent them as a composition of moving objects. Recent works model a set of 2D sprites over a slowly-varying background, but without considering the underlying 3D scene that gives rise to them. We instead propose to model a video as the view seen while moving through a scene with multiple 3D objects and a 3D background. Our model is trained from monocular videos without any supervision, yet learns to generate coherent 3D scenes containing several moving objects. We conduct detailed experiments on two datasets, going beyond the visual complexity supported by state-of-the-art generative approaches. We evaluate our method on depth-prediction and 3D object detection---tasks which cannot be addressed by those earlier works---and show it out-performs them even on 2D instance segmentation and tracking. AU - Henderson, Paul M AU - Lampert, Christoph ID - 8188 SN - 9781713829546 T2 - 34th Conference on Neural Information Processing Systems TI - Unsupervised object-centric video generation and decomposition in 3D VL - 33 ER - TY - JOUR AB - We present a unified framework tackling two problems: class-specific 3D reconstruction from a single image, and generation of new 3D shape samples. These tasks have received considerable attention recently; however, most existing approaches rely on 3D supervision, annotation of 2D images with keypoints or poses, and/or training with multiple views of each object instance. Our framework is very general: it can be trained in similar settings to existing approaches, while also supporting weaker supervision. Importantly, it can be trained purely from 2D images, without pose annotations, and with only a single view per instance. We employ meshes as an output representation, instead of voxels used in most prior work. This allows us to reason over lighting parameters and exploit shading information during training, which previous 2D-supervised methods cannot. Thus, our method can learn to generate and reconstruct concave object classes. We evaluate our approach in various settings, showing that: (i) it learns to disentangle shape from pose and lighting; (ii) using shading in the loss improves performance compared to just silhouettes; (iii) when using a standard single white light, our model outperforms state-of-the-art 2D-supervised methods, both with and without pose supervision, thanks to exploiting shading cues; (iv) performance improves further when using multiple coloured lights, even approaching that of state-of-the-art 3D-supervised methods; (v) shapes produced by our model capture smooth surfaces and fine details better than voxel-based approaches; and (vi) our approach supports concave classes such as bathtubs and sofas, which methods based on silhouettes cannot learn. AU - Henderson, Paul M AU - Ferrari, Vittorio ID - 6952 JF - International Journal of Computer Vision SN - 0920-5691 TI - Learning single-image 3D reconstruction by generative modelling of shape, pose and shading VL - 128 ER - TY - CONF AB - State-of-the-art detection systems are generally evaluated on their ability to exhaustively retrieve objects densely distributed in the image, across a wide variety of appearances and semantic categories. Orthogonal to this, many real-life object detection applications, for example in remote sensing, instead require dealing with large images that contain only a few small objects of a single class, scattered heterogeneously across the space. In addition, they are often subject to strict computational constraints, such as limited battery capacity and computing power.To tackle these more practical scenarios, we propose a novel flexible detection scheme that efficiently adapts to variable object sizes and densities: We rely on a sequence of detection stages, each of which has the ability to predict groups of objects as well as individuals. Similar to a detection cascade, this multi-stage architecture spares computational effort by discarding large irrelevant regions of the image early during the detection process. The ability to group objects provides further computational and memory savings, as it allows working with lower image resolutions in early stages, where groups are more easily detected than individuals, as they are more salient. We report experimental results on two aerial image datasets, and show that the proposed method is as accurate yet computationally more efficient than standard single-shot detectors, consistently across three different backbone architectures. AU - Royer, Amélie AU - Lampert, Christoph ID - 7936 SN - 9781728165530 T2 - IEEE Winter Conference on Applications of Computer Vision TI - Localizing grouped instances for efficient detection in low-resource scenarios ER - TY - CONF AB - Fine-tuning is a popular way of exploiting knowledge contained in a pre-trained convolutional network for a new visual recognition task. However, the orthogonal setting of transferring knowledge from a pretrained network to a visually different yet semantically close source is rarely considered: This commonly happens with real-life data, which is not necessarily as clean as the training source (noise, geometric transformations, different modalities, etc.).To tackle such scenarios, we introduce a new, generalized form of fine-tuning, called flex-tuning, in which any individual unit (e.g. layer) of a network can be tuned, and the most promising one is chosen automatically. In order to make the method appealing for practical use, we propose two lightweight and faster selection procedures that prove to be good approximations in practice. We study these selection criteria empirically across a variety of domain shifts and data scarcity scenarios, and show that fine-tuning individual units, despite its simplicity, yields very good results as an adaptation technique. As it turns out, in contrast to common practice, rather than the last fully-connected unit it is best to tune an intermediate or early one in many domain- shift scenarios, which is accurately detected by flex-tuning. AU - Royer, Amélie AU - Lampert, Christoph ID - 7937 SN - 9781728165530 T2 - 2020 IEEE Winter Conference on Applications of Computer Vision TI - A flexible selection scheme for minimum-effort transfer learning ER - TY - CHAP AB - Image translation refers to the task of mapping images from a visual domain to another. Given two unpaired collections of images, we aim to learn a mapping between the corpus-level style of each collection, while preserving semantic content shared across the two domains. We introduce xgan, a dual adversarial auto-encoder, which captures a shared representation of the common domain semantic content in an unsupervised way, while jointly learning the domain-to-domain image translations in both directions. We exploit ideas from the domain adaptation literature and define a semantic consistency loss which encourages the learned embedding to preserve semantics shared across domains. We report promising qualitative results for the task of face-to-cartoon translation. The cartoon dataset we collected for this purpose, “CartoonSet”, is also publicly available as a new benchmark for semantic style transfer at https://google.github.io/cartoonset/index.html. AU - Royer, Amélie AU - Bousmalis, Konstantinos AU - Gouws, Stephan AU - Bertsch, Fred AU - Mosseri, Inbar AU - Cole, Forrester AU - Murphy, Kevin ED - Singh, Richa ED - Vatsa, Mayank ED - Patel, Vishal M. ED - Ratha, Nalini ID - 8092 SN - 9783030306717 T2 - Domain Adaptation for Visual Understanding TI - XGAN: Unsupervised image-to-image translation for many-to-many mappings ER - TY - CONF AB - We address the following question: How redundant is the parameterisation of ReLU networks? Specifically, we consider transformations of the weight space which leave the function implemented by the network intact. Two such transformations are known for feed-forward architectures: permutation of neurons within a layer, and positive scaling of all incoming weights of a neuron coupled with inverse scaling of its outgoing weights. In this work, we show for architectures with non-increasing widths that permutation and scaling are in fact the only function-preserving weight transformations. For any eligible architecture we give an explicit construction of a neural network such that any other network that implements the same function can be obtained from the original one by the application of permutations and rescaling. The proof relies on a geometric understanding of boundaries between linear regions of ReLU networks, and we hope the developed mathematical tools are of independent interest. AU - Bui Thi Mai, Phuong AU - Lampert, Christoph ID - 7481 T2 - 8th International Conference on Learning Representations TI - Functional vs. parametric equivalence of ReLU networks ER - TY - CONF AB - We study the problem of learning from multiple untrusted data sources, a scenario of increasing practical relevance given the recent emergence of crowdsourcing and collaborative learning paradigms. Specifically, we analyze the situation in which a learning system obtains datasets from multiple sources, some of which might be biased or even adversarially perturbed. It is known that in the single-source case, an adversary with the power to corrupt a fixed fraction of the training data can prevent PAC-learnability, that is, even in the limit of infinitely much training data, no learning system can approach the optimal test error. In this work we show that, surprisingly, the same is not true in the multi-source setting, where the adversary can arbitrarily corrupt a fixed fraction of the data sources. Our main results are a generalization bound that provides finite-sample guarantees for this learning setting, as well as corresponding lower bounds. Besides establishing PAC-learnability our results also show that in a cooperative learning setting sharing data with other parties has provable benefits, even if some participants are malicious. AU - Konstantinov, Nikola H AU - Frantar, Elias AU - Alistarh, Dan-Adrian AU - Lampert, Christoph ID - 8724 SN - 2640-3498 T2 - Proceedings of the 37th International Conference on Machine Learning TI - On the sample complexity of adversarial multi-source PAC learning VL - 119 ER - TY - THES AB - Deep neural networks have established a new standard for data-dependent feature extraction pipelines in the Computer Vision literature. Despite their remarkable performance in the standard supervised learning scenario, i.e. when models are trained with labeled data and tested on samples that follow a similar distribution, neural networks have been shown to struggle with more advanced generalization abilities, such as transferring knowledge across visually different domains, or generalizing to new unseen combinations of known concepts. In this thesis we argue that, in contrast to the usual black-box behavior of neural networks, leveraging more structured internal representations is a promising direction for tackling such problems. In particular, we focus on two forms of structure. First, we tackle modularity: We show that (i) compositional architectures are a natural tool for modeling reasoning tasks, in that they efficiently capture their combinatorial nature, which is key for generalizing beyond the compositions seen during training. We investigate how to to learn such models, both formally and experimentally, for the task of abstract visual reasoning. Then, we show that (ii) in some settings, modularity allows us to efficiently break down complex tasks into smaller, easier, modules, thereby improving computational efficiency; We study this behavior in the context of generative models for colorization, as well as for small objects detection. Secondly, we investigate the inherently layered structure of representations learned by neural networks, and analyze its role in the context of transfer learning and domain adaptation across visually dissimilar domains. AU - Royer, Amélie ID - 8390 SN - 2663-337X TI - Leveraging structure in Computer Vision tasks for flexible Deep Learning models ER - TY - CONF AB - Numerous methods have been proposed for probabilistic generative modelling of 3D objects. However, none of these is able to produce textured objects, which renders them of limited use for practical tasks. In this work, we present the first generative model of textured 3D meshes. Training such a model would traditionally require a large dataset of textured meshes, but unfortunately, existing datasets of meshes lack detailed textures. We instead propose a new training methodology that allows learning from collections of 2D images without any 3D information. To do so, we train our model to explain a distribution of images by modelling each image as a 3D foreground object placed in front of a 2D background. Thus, it learns to generate meshes that when rendered, produce images similar to those in its training set. A well-known problem when generating meshes with deep networks is the emergence of self-intersections, which are problematic for many use-cases. As a second contribution we therefore introduce a new generation process for 3D meshes that guarantees no self-intersections arise, based on the physical intuition that faces should push one another out of the way as they move. We conduct extensive experiments on our approach, reporting quantitative and qualitative results on both synthetic data and natural images. These show our method successfully learns to generate plausible and diverse textured 3D samples for five challenging object classes. AU - Henderson, Paul M AU - Tsiminaki, Vagia AU - Lampert, Christoph ID - 8186 T2 - Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition TI - Leveraging 2D data to learn textured 3D mesh generation ER - TY - JOUR AB - We study the problem of automatically detecting if a given multi-class classifier operates outside of its specifications (out-of-specs), i.e. on input data from a different distribution than what it was trained for. This is an important problem to solve on the road towards creating reliable computer vision systems for real-world applications, because the quality of a classifier’s predictions cannot be guaranteed if it operates out-of-specs. Previously proposed methods for out-of-specs detection make decisions on the level of single inputs. This, however, is insufficient to achieve low false positive rate and high false negative rates at the same time. In this work, we describe a new procedure named KS(conf), based on statistical reasoning. Its main component is a classical Kolmogorov–Smirnov test that is applied to the set of predicted confidence values for batches of samples. Working with batches instead of single samples allows increasing the true positive rate without negatively affecting the false positive rate, thereby overcoming a crucial limitation of single sample tests. We show by extensive experiments using a variety of convolutional network architectures and datasets that KS(conf) reliably detects out-of-specs situations even under conditions where other tests fail. It furthermore has a number of properties that make it an excellent candidate for practical deployment: it is easy to implement, adds almost no overhead to the system, works with any classifier that outputs confidence scores, and requires no a priori knowledge about how the data distribution could change. AU - Sun, Rémy AU - Lampert, Christoph ID - 6944 IS - 4 JF - International Journal of Computer Vision SN - 0920-5691 TI - KS(conf): A light-weight test if a multiclass classifier operates outside of its specifications VL - 128 ER - TY - BOOK AB - Wissen Sie, was sich hinter künstlicher Intelligenz und maschinellem Lernen verbirgt? Dieses Sachbuch erklärt Ihnen leicht verständlich und ohne komplizierte Formeln die grundlegenden Methoden und Vorgehensweisen des maschinellen Lernens. Mathematisches Vorwissen ist dafür nicht nötig. Kurzweilig und informativ illustriert Lisa, die Protagonistin des Buches, diese anhand von Alltagssituationen. Ein Buch für alle, die in Diskussionen über Chancen und Risiken der aktuellen Entwicklung der künstlichen Intelligenz und des maschinellen Lernens mit Faktenwissen punkten möchten. Auch für Schülerinnen und Schüler geeignet! ED - Kersting, Kristian ED - Lampert, Christoph ED - Rothkopf, Constantin ID - 7171 SN - 978-3-658-26762-9 TI - Wie Maschinen Lernen: Künstliche Intelligenz Verständlich Erklärt ER - TY - CONF AB - Graph games and Markov decision processes (MDPs) are standard models in reactive synthesis and verification of probabilistic systems with nondeterminism. The class of 𝜔 -regular winning conditions; e.g., safety, reachability, liveness, parity conditions; provides a robust and expressive specification formalism for properties that arise in analysis of reactive systems. The resolutions of nondeterminism in games and MDPs are represented as strategies, and we consider succinct representation of such strategies. The decision-tree data structure from machine learning retains the flavor of decisions of strategies and allows entropy-based minimization to obtain succinct trees. However, in contrast to traditional machine-learning problems where small errors are allowed, for winning strategies in graph games and MDPs no error is allowed, and the decision tree must represent the entire strategy. In this work we propose decision trees with linear classifiers for representation of strategies in graph games and MDPs. We have implemented strategy representation using this data structure and we present experimental results for problems on graph games and MDPs, which show that this new data structure presents a much more efficient strategy representation as compared to standard decision trees. AU - Ashok, Pranav AU - Brázdil, Tomáš AU - Chatterjee, Krishnendu AU - Křetínský, Jan AU - Lampert, Christoph AU - Toman, Viktor ID - 6942 SN - 0302-9743 T2 - 16th International Conference on Quantitative Evaluation of Systems TI - Strategy representation by decision trees with linear classifiers VL - 11785 ER - TY - JOUR AB - Due to the importance of zero-shot learning, i.e. classifying images where there is a lack of labeled training data, the number of proposed approaches has recently increased steadily. We argue that it is time to take a step back and to analyze the status quo of the area. The purpose of this paper is three-fold. First, given the fact that there is no agreed upon zero-shot learning benchmark, we first define a new benchmark by unifying both the evaluation protocols and data splits of publicly available datasets used for this task. This is an important contribution as published results are often not comparable and sometimes even flawed due to, e.g. pre-training on zero-shot test classes. Moreover, we propose a new zero-shot learning dataset, the Animals with Attributes 2 (AWA2) dataset which we make publicly available both in terms of image features and the images themselves. Second, we compare and analyze a significant number of the state-of-the-art methods in depth, both in the classic zero-shot setting but also in the more realistic generalized zero-shot setting. Finally, we discuss in detail the limitations of the current status of the area which can be taken as a basis for advancing it. AU - Xian, Yongqin AU - Lampert, Christoph AU - Schiele, Bernt AU - Akata, Zeynep ID - 6554 IS - 9 JF - IEEE Transactions on Pattern Analysis and Machine Intelligence SN - 0162-8828 TI - Zero-shot learning - A comprehensive evaluation of the good, the bad and the ugly VL - 41 ER - TY - CONF AB - Multi-exit architectures, in which a stack of processing layers is interleaved with early output layers, allow the processing of a test example to stop early and thus save computation time and/or energy. In this work, we propose a new training procedure for multi-exit architectures based on the principle of knowledge distillation. The method encourage searly exits to mimic later, more accurate exits, by matching their output probabilities. Experiments on CIFAR100 and ImageNet show that distillation-based training significantly improves the accuracy of early exits while maintaining state-of-the-art accuracy for late ones. The method is particularly beneficial when training data is limited and it allows a straightforward extension to semi-supervised learning,i.e. making use of unlabeled data at training time. Moreover, it takes only afew lines to implement and incurs almost no computational overhead at training time, and none at all at test time. AU - Bui Thi Mai, Phuong AU - Lampert, Christoph ID - 7479 SN - 15505499 T2 - IEEE International Conference on Computer Vision TI - Distillation-based training for multi-exit architectures VL - 2019-October ER - TY - CONF AB - We propose a new model for detecting visual relationships, such as "person riding motorcycle" or "bottle on table". This task is an important step towards comprehensive structured mage understanding, going beyond detecting individual objects. Our main novelty is a Box Attention mechanism that allows to model pairwise interactions between objects using standard object detection pipelines. The resulting model is conceptually clean, expressive and relies on well-justified training and prediction procedures. Moreover, unlike previously proposed approaches, our model does not introduce any additional complex components or hyperparameters on top of those already required by the underlying detection model. We conduct an experimental evaluation on two datasets, V-COCO and Open Images, demonstrating strong quantitative and qualitative results. AU - Kolesnikov, Alexander AU - Kuznetsova, Alina AU - Lampert, Christoph AU - Ferrari, Vittorio ID - 7640 SN - 9781728150239 T2 - Proceedings of the 2019 International Conference on Computer Vision Workshop TI - Detecting visual relationships using box attention ER - TY - CONF AB - Knowledge distillation, i.e. one classifier being trained on the outputs of another classifier, is an empirically very successful technique for knowledge transfer between classifiers. It has even been observed that classifiers learn much faster and more reliably if trained with the outputs of another classifier as soft labels, instead of from ground truth data. So far, however, there is no satisfactory theoretical explanation of this phenomenon. In this work, we provide the first insights into the working mechanisms of distillation by studying the special case of linear and deep linear classifiers. Specifically, we prove a generalization bound that establishes fast convergence of the expected risk of a distillation-trained linear classifier. From the bound and its proof we extract three keyfactors that determine the success of distillation: data geometry – geometric properties of the datadistribution, in particular class separation, has an immediate influence on the convergence speed of the risk; optimization bias– gradient descentoptimization finds a very favorable minimum of the distillation objective; and strong monotonicity– the expected risk of the student classifier always decreases when the size of the training set grows. AU - Bui Thi Mai, Phuong AU - Lampert, Christoph ID - 6569 T2 - Proceedings of the 36th International Conference on Machine Learning TI - Towards understanding knowledge distillation VL - 97 ER - TY - CONF AB - Modern machine learning methods often require more data for training than a single expert can provide. Therefore, it has become a standard procedure to collect data from external sources, e.g. via crowdsourcing. Unfortunately, the quality of these sources is not always guaranteed. As additional complications, the data might be stored in a distributed way, or might even have to remain private. In this work, we address the question of how to learn robustly in such scenarios. Studying the problem through the lens of statistical learning theory, we derive a procedure that allows for learning from all available sources, yet automatically suppresses irrelevant or corrupted data. We show by extensive experiments that our method provides significant improvements over alternative approaches from robust statistics and distributed optimization. AU - Konstantinov, Nikola H AU - Lampert, Christoph ID - 6590 T2 - Proceedings of the 36th International Conference on Machine Learning TI - Robust learning from untrusted sources VL - 97 ER - TY - CONF AB - Computer vision systems for automatic image categorization have become accurate and reliable enough that they can run continuously for days or even years as components of real-world commercial applications. A major open problem in this context, however, is quality control. Good classification performance can only be expected if systems run under the specific conditions, in particular data distributions, that they were trained for. Surprisingly, none of the currently used deep network architectures have a built-in functionality that could detect if a network operates on data from a distribution it was not trained for, such that potentially a warning to the human users could be triggered. In this work, we describe KS(conf), a procedure for detecting such outside of specifications (out-of-specs) operation, based on statistical testing of the network outputs. We show by extensive experiments using the ImageNet, AwA2 and DAVIS datasets on a variety of ConvNets architectures that KS(conf) reliably detects out-of-specs situations. It furthermore has a number of properties that make it a promising candidate for practical deployment: it is easy to implement, adds almost no overhead to the system, works with all networks, including pretrained ones, and requires no a priori knowledge of how the data distribution could change. AU - Sun, Rémy AU - Lampert, Christoph ID - 6482 SN - 0302-9743 TI - KS(conf): A light-weight test if a ConvNet operates outside of Its specifications VL - 11269 ER - TY - THES AB - The most common assumption made in statistical learning theory is the assumption of the independent and identically distributed (i.i.d.) data. While being very convenient mathematically, it is often very clearly violated in practice. This disparity between the machine learning theory and applications underlies a growing demand in the development of algorithms that learn from dependent data and theory that can provide generalization guarantees similar to the independent situations. This thesis is dedicated to two variants of dependencies that can arise in practice. One is a dependence on the level of samples in a single learning task. Another dependency type arises in the multi-task setting when the tasks are dependent on each other even though the data for them can be i.i.d. In both cases we model the data (samples or tasks) as stochastic processes and introduce new algorithms for both settings that take into account and exploit the resulting dependencies. We prove the theoretical guarantees on the performance of the introduced algorithms under different evaluation criteria and, in addition, we compliment the theoretical study by the empirical one, where we evaluate some of the algorithms on two real world datasets to highlight their practical applicability. AU - Zimin, Alexander ID - 68 SN - 2663-337X TI - Learning from dependent data ER - TY - THES AB - Modern computer vision systems heavily rely on statistical machine learning models, which typically require large amounts of labeled data to be learned reliably. Moreover, very recently computer vision research widely adopted techniques for representation learning, which further increase the demand for labeled data. However, for many important practical problems there is relatively small amount of labeled data available, so it is problematic to leverage full potential of the representation learning methods. One way to overcome this obstacle is to invest substantial resources into producing large labelled datasets. Unfortunately, this can be prohibitively expensive in practice. In this thesis we focus on the alternative way of tackling the aforementioned issue. We concentrate on methods, which make use of weakly-labeled or even unlabeled data. Specifically, the first half of the thesis is dedicated to the semantic image segmentation task. We develop a technique, which achieves competitive segmentation performance and only requires annotations in a form of global image-level labels instead of dense segmentation masks. Subsequently, we present a new methodology, which further improves segmentation performance by leveraging tiny additional feedback from a human annotator. By using our methods practitioners can greatly reduce the amount of data annotation effort, which is required to learn modern image segmentation models. In the second half of the thesis we focus on methods for learning from unlabeled visual data. We study a family of autoregressive models for modeling structure of natural images and discuss potential applications of these models. Moreover, we conduct in-depth study of one of these applications, where we develop the state-of-the-art model for the probabilistic image colorization task. AU - Kolesnikov, Alexander ID - 197 SN - 2663-337X TI - Weakly-Supervised Segmentation and Unsupervised Modeling of Natural Images ER - TY - JOUR AB - In continuous populations with local migration, nearby pairs of individuals have on average more similar genotypes than geographically well separated pairs. A barrier to gene flow distorts this classical pattern of isolation by distance. Genetic similarity is decreased for sample pairs on different sides of the barrier and increased for pairs on the same side near the barrier. Here, we introduce an inference scheme that utilizes this signal to detect and estimate the strength of a linear barrier to gene flow in two-dimensions. We use a diffusion approximation to model the effects of a barrier on the geographical spread of ancestry backwards in time. This approach allows us to calculate the chance of recent coalescence and probability of identity by descent. We introduce an inference scheme that fits these theoretical results to the geographical covariance structure of bialleleic genetic markers. It can estimate the strength of the barrier as well as several demographic parameters. We investigate the power of our inference scheme to detect barriers by applying it to a wide range of simulated data. We also showcase an example application to a Antirrhinum majus (snapdragon) flower color hybrid zone, where we do not detect any signal of a strong genome wide barrier to gene flow. AU - Ringbauer, Harald AU - Kolesnikov, Alexander AU - Field, David AU - Barton, Nicholas H ID - 563 IS - 3 JF - Genetics TI - Estimating barriers to gene flow from distorted isolation-by-distance patterns VL - 208 ER - TY - JOUR AB - The twelve papers in this special section focus on learning systems with shared information for computer vision and multimedia communication analysis. In the real world, a realistic setting for computer vision or multimedia recognition problems is that we have some classes containing lots of training data and many classes containing a small amount of training data. Therefore, how to use frequent classes to help learning rare classes for which it is harder to collect the training data is an open question. Learning with shared information is an emerging topic in machine learning, computer vision and multimedia analysis. There are different levels of components that can be shared during concept modeling and machine learning stages, such as sharing generic object parts, sharing attributes, sharing transformations, sharing regularization parameters and sharing training examples, etc. Regarding the specific methods, multi-task learning, transfer learning and deep learning can be seen as using different strategies to share information. These learning with shared information methods are very effective in solving real-world large-scale problems. AU - Darrell, Trevor AU - Lampert, Christoph AU - Sebe, Nico AU - Wu, Ying AU - Yan, Yan ID - 321 IS - 5 JF - IEEE Transactions on Pattern Analysis and Machine Intelligence TI - Guest editors' introduction to the special section on learning with Shared information for computer vision and multimedia analysis VL - 40 ER - TY - CONF AB - We introduce Intelligent Annotation Dialogs for bounding box annotation. We train an agent to automatically choose a sequence of actions for a human annotator to produce a bounding box in a minimal amount of time. Specifically, we consider two actions: box verification [34], where the annotator verifies a box generated by an object detector, and manual box drawing. We explore two kinds of agents, one based on predicting the probability that a box will be positively verified, and the other based on reinforcement learning. We demonstrate that (1) our agents are able to learn efficient annotation strategies in several scenarios, automatically adapting to the image difficulty, the desired quality of the boxes, and the detector strength; (2) in all scenarios the resulting annotation dialogs speed up annotation compared to manual box drawing alone and box verification alone, while also outperforming any fixed combination of verification and drawing in most scenarios; (3) in a realistic scenario where the detector is iteratively re-trained, our agents evolve a series of strategies that reflect the shifting trade-off between verification and drawing as the detector grows stronger. AU - Uijlings, Jasper AU - Konyushkova, Ksenia AU - Lampert, Christoph AU - Ferrari, Vittorio ID - 10882 SN - 9781538664209 T2 - 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition TI - Learning intelligent dialogs for bounding box annotation ER - TY - CONF AB - We present an approach to identify concise equations from data using a shallow neural network approach. In contrast to ordinary black-box regression, this approach allows understanding functional relations and generalizing them from observed data to unseen parts of the parameter space. We show how to extend the class of learnable equations for a recently proposed equation learning network to include divisions, and we improve the learning and model selection strategy to be useful for challenging real-world data. For systems governed by analytical expressions, our method can in many cases identify the true underlying equation and extrapolate to unseen domains. We demonstrate its effectiveness by experiments on a cart-pendulum system, where only 2 random rollouts are required to learn the forward dynamics and successfully achieve the swing-up task. AU - Sahoo, Subham AU - Lampert, Christoph AU - Martius, Georg S ID - 6012 T2 - Proceedings of the 35th International Conference on Machine Learning TI - Learning equations for extrapolation and control VL - 80 ER - TY - CONF AB - We establish a data-dependent notion of algorithmic stability for Stochastic Gradient Descent (SGD), and employ it to develop novel generalization bounds. This is in contrast to previous distribution-free algorithmic stability results for SGD which depend on the worst-case constants. By virtue of the data-dependent argument, our bounds provide new insights into learning with SGD on convex and non-convex problems. In the convex case, we show that the bound on the generalization error depends on the risk at the initialization point. In the non-convex case, we prove that the expected curvature of the objective function around the initialization point has crucial influence on the generalization error. In both cases, our results suggest a simple data-driven strategy to stabilize SGD by pre-screening its initialization. As a corollary, our results allow us to show optimistic generalization bounds that exhibit fast convergence rates for SGD subject to a vanishing empirical risk and low noise of stochastic gradient. AU - Kuzborskij, Ilja AU - Lampert, Christoph ID - 6011 T2 - Proceedings of the 35 th International Conference on Machine Learning TI - Data-dependent stability of stochastic gradient descent VL - 80 ER - TY - CONF AB - Distributed training of massive machine learning models, in particular deep neural networks, via Stochastic Gradient Descent (SGD) is becoming commonplace. Several families of communication-reduction methods, such as quantization, large-batch methods, and gradient sparsification, have been proposed. To date, gradient sparsification methods--where each node sorts gradients by magnitude, and only communicates a subset of the components, accumulating the rest locally--are known to yield some of the largest practical gains. Such methods can reduce the amount of communication per step by up to \emph{three orders of magnitude}, while preserving model accuracy. Yet, this family of methods currently has no theoretical justification. This is the question we address in this paper. We prove that, under analytic assumptions, sparsifying gradients by magnitude with local error correction provides convergence guarantees, for both convex and non-convex smooth objectives, for data-parallel SGD. The main insight is that sparsification methods implicitly maintain bounds on the maximum impact of stale updates, thanks to selection by magnitude. Our analysis and empirical validation also reveal that these methods do require analytical conditions to converge well, justifying existing heuristics. AU - Alistarh, Dan-Adrian AU - Hoefler, Torsten AU - Johansson, Mikael AU - Konstantinov, Nikola H AU - Khirirat, Sarit AU - Renggli, Cedric ID - 6589 T2 - Advances in Neural Information Processing Systems 31 TI - The convergence of sparsified gradient methods VL - Volume 2018 ER -