TY - CONF
AB - We address the problem of metric learning for multi-view data, namely the construction of embedding projections from data in different representations into a shared feature space, such that the Euclidean distance in this space provides a meaningful within-view as well as between-view similarity. Our motivation stems from the problem of cross-media retrieval tasks, where the availability of a joint Euclidean distance function is a pre-requisite to allow fast, in particular hashing-based, nearest neighbor queries. We formulate an objective function that expresses the intuitive concept that matching samples are mapped closely together in the output space, whereas non-matching samples are pushed apart, no matter in which view they are available. The resulting optimization problem is not convex, but it can be decomposed explicitly into a convex and a concave part, thereby allowing efficient optimization using the convex-concave procedure. Experiments on an image retrieval task show that nearest-neighbor based cross-view retrieval is indeed possible, and the proposed technique improves the retrieval accuracy over baseline techniques.
AU - Quadrianto, Novi
AU - Lampert, Christoph
ID - 3319
TI - Learning multi-view neighborhood preserving projections
ER -
TY - JOUR
AB - Powerful statistical models that can be learned efficiently from large amounts of data are currently revolutionizing computer vision. These models possess a rich internal structure reflecting task-specific relations and constraints. This monograph introduces the reader to the most popular classes of structured models in computer vision. Our focus is discrete undirected graphical models which we cover in detail together with a description of algorithms for both probabilistic inference and maximum a posteriori inference. We discuss separately recently successful techniques for prediction in general structured models. In the second part of this monograph we describe methods for parameter learning where we distinguish the classic maximum likelihood based methods from the more recent prediction-based parameter learning methods. We highlight developments to enhance current models and discuss kernelized models and latent variable models. To make the monograph more practical and to provide links to further study we provide examples of successful application of many methods in the computer vision literature.
AU - Nowozin, Sebastian
AU - Lampert, Christoph
ID - 3320
IS - 3-4
JF - Foundations and Trends in Computer Graphics and Vision
TI - Structured learning and prediction in computer vision
VL - 6
ER -
TY - GEN
AB - We study multi-label prediction for structured output spaces, a problem that occurs, for example, in object detection in images, secondary structure prediction in computational biology, and graph matching with symmetries. Conventional multi-label classification techniques are typically not applicable in this situation, because they require explicit enumeration of the label space, which is infeasible in case of structured outputs. Relying on techniques originally designed for single- label structured prediction, in particular structured support vector machines, results in reduced prediction accuracy, or leads to infeasible optimization problems. In this work we derive a maximum-margin training formulation for multi-label structured prediction that remains computationally tractable while achieving high prediction accuracy. It also shares most beneficial properties with single-label maximum-margin approaches, in particular a formulation as a convex optimization problem, efficient working set training, and PAC-Bayesian generalization bounds.
AU - Lampert, Christoph
ID - 3322
T2 - NIPS: Neural Information Processing Systems
TI - Maximum margin multi label structured prediction
ER -
TY - CONF
AB - We present a new decidable logic called TREX for expressing constraints about imperative tree data structures. In particular, TREX supports a transitive closure operator that can express reachability constraints, which often appear in data structure invariants. We show that our logic is closed under weakest precondition computation, which enables its use for automated software verification. We further show that satisfiability of formulas in TREX is decidable in NP. The low complexity makes it an attractive alternative to more expensive logics such as monadic second-order logic (MSOL) over trees, which have been traditionally used for reasoning about tree data structures.
AU - Wies, Thomas
AU - MuĂ±iz, Marco
AU - Kuncak, Viktor
ID - 3323
TI - An efficient decision procedure for imperative tree data structures
VL - 6803
ER -
TY - CONF
AB - Automated termination provers often use the following schema to prove that a program terminates: construct a relational abstraction of the program's transition relation and then show that the relational abstraction is well-founded. The focus of current tools has been on developing sophisticated techniques for constructing the abstractions while relying on known decidable logics (such as linear arithmetic) to express them. We believe we can significantly increase the class of programs that are amenable to automated termination proofs by identifying more expressive decidable logics for reasoning about well-founded relations. We therefore present a new decision procedure for reasoning about multiset orderings, which are among the most powerful orderings used to prove termination. We show that, using our decision procedure, one can automatically prove termination of natural abstractions of programs.
AU - Piskac, Ruzica
AU - Wies, Thomas
ED - Jhala, Ranjit
ED - Schmidt, David
ID - 3324
TI - Decision procedures for automating termination proofs
VL - 6538
ER -