--- _id: '2189' abstract: - lang: fre text: En apprentissage automatique, nous parlons d'adaptation de domaine lorsque les données de test (cibles) et d'apprentissage (sources) sont générées selon différentes distributions. Nous devons donc développer des algorithmes de classification capables de s'adapter à une nouvelle distribution, pour laquelle aucune information sur les étiquettes n'est disponible. Nous attaquons cette problématique sous l'angle de l'approche PAC-Bayésienne qui se focalise sur l'apprentissage de modèles définis comme des votes de majorité sur un ensemble de fonctions. Dans ce contexte, nous introduisons PV-MinCq une version adaptative de l'algorithme (non adaptatif) MinCq. PV-MinCq suit le principe suivant. Nous transférons les étiquettes sources aux points cibles proches pour ensuite appliquer MinCq sur l'échantillon cible ``auto-étiqueté'' (justifié par une borne théorique). Plus précisément, nous définissons un auto-étiquetage non itératif qui se focalise dans les régions où les distributions marginales source et cible sont les plus similaires. Dans un second temps, nous étudions l'influence de notre auto-étiquetage pour en déduire une procédure de validation des hyperparamètres. Finalement, notre approche montre des résultats empiriques prometteurs. article_processing_charge: No author: - first_name: Emilie full_name: Morvant, Emilie id: 4BAC2A72-F248-11E8-B48F-1D18A9856A87 last_name: Morvant orcid: 0000-0002-8301-7240 citation: ama: 'Morvant E. Adaptation de domaine de vote de majorité par auto-étiquetage non itératif. In: Vol 1. Elsevier; 2014:49-58.' apa: 'Morvant, E. (2014). Adaptation de domaine de vote de majorité par auto-étiquetage non itératif (Vol. 1, pp. 49–58). Presented at the CAP: Conférence Francophone sur l’Apprentissage Automatique (Machine Learning French Conference), Saint-Etienne, France: Elsevier.' chicago: Morvant, Emilie. “Adaptation de Domaine de Vote de Majorité Par Auto-Étiquetage Non Itératif,” 1:49–58. Elsevier, 2014. ieee: 'E. Morvant, “Adaptation de domaine de vote de majorité par auto-étiquetage non itératif,” presented at the CAP: Conférence Francophone sur l’Apprentissage Automatique (Machine Learning French Conference), Saint-Etienne, France, 2014, vol. 1, pp. 49–58.' ista: 'Morvant E. 2014. Adaptation de domaine de vote de majorité par auto-étiquetage non itératif. CAP: Conférence Francophone sur l’Apprentissage Automatique (Machine Learning French Conference) vol. 1, 49–58.' mla: Morvant, Emilie. Adaptation de Domaine de Vote de Majorité Par Auto-Étiquetage Non Itératif. Vol. 1, Elsevier, 2014, pp. 49–58. short: E. Morvant, in:, Elsevier, 2014, pp. 49–58. conference: location: Saint-Etienne, France name: 'CAP: Conférence Francophone sur l''Apprentissage Automatique (Machine Learning French Conference)' date_created: 2018-12-11T11:56:13Z date_published: 2014-07-01T00:00:00Z date_updated: 2021-01-12T06:55:52Z day: '01' department: - _id: ChLa intvolume: ' 1' language: - iso: eng main_file_link: - open_access: '1' url: https://hal.archives-ouvertes.fr/hal-01005776/ month: '07' oa: 1 oa_version: Preprint page: 49-58 publication_status: published publisher: Elsevier publist_id: '4785' quality_controlled: '1' status: public title: Adaptation de domaine de vote de majorité par auto-étiquetage non itératif type: conference user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87 volume: 1 year: '2014' ... --- _id: '2160' abstract: - lang: eng text: Transfer learning has received a lot of attention in the machine learning community over the last years, and several effective algorithms have been developed. However, relatively little is known about their theoretical properties, especially in the setting of lifelong learning, where the goal is to transfer information to tasks for which no data have been observed so far. In this work we study lifelong learning from a theoretical perspective. Our main result is a PAC-Bayesian generalization bound that offers a unified view on existing paradigms for transfer learning, such as the transfer of parameters or the transfer of low-dimensional representations. We also use the bound to derive two principled lifelong learning algorithms, and we show that these yield results comparable with existing methods. article_processing_charge: No author: - first_name: Anastasia full_name: Pentina, Anastasia id: 42E87FC6-F248-11E8-B48F-1D18A9856A87 last_name: Pentina - first_name: Christoph full_name: Lampert, Christoph id: 40C20FD2-F248-11E8-B48F-1D18A9856A87 last_name: Lampert orcid: 0000-0001-8622-7887 citation: ama: 'Pentina A, Lampert C. A PAC-Bayesian bound for Lifelong Learning. In: Vol 32. ML Research Press; 2014:991-999.' apa: 'Pentina, A., & Lampert, C. (2014). A PAC-Bayesian bound for Lifelong Learning (Vol. 32, pp. 991–999). Presented at the ICML: International Conference on Machine Learning, Beijing, China: ML Research Press.' chicago: Pentina, Anastasia, and Christoph Lampert. “A PAC-Bayesian Bound for Lifelong Learning,” 32:991–99. ML Research Press, 2014. ieee: 'A. Pentina and C. Lampert, “A PAC-Bayesian bound for Lifelong Learning,” presented at the ICML: International Conference on Machine Learning, Beijing, China, 2014, vol. 32, pp. 991–999.' ista: 'Pentina A, Lampert C. 2014. A PAC-Bayesian bound for Lifelong Learning. ICML: International Conference on Machine Learning vol. 32, 991–999.' mla: Pentina, Anastasia, and Christoph Lampert. A PAC-Bayesian Bound for Lifelong Learning. Vol. 32, ML Research Press, 2014, pp. 991–99. short: A. Pentina, C. Lampert, in:, ML Research Press, 2014, pp. 991–999. conference: end_date: 2014-06-26 location: Beijing, China name: 'ICML: International Conference on Machine Learning' start_date: 2014-06-21 date_created: 2018-12-11T11:56:03Z date_published: 2014-05-10T00:00:00Z date_updated: 2023-10-17T11:54:24Z day: '10' department: - _id: ChLa intvolume: ' 32' language: - iso: eng main_file_link: - open_access: '1' url: https://dl.acm.org/citation.cfm?id=3045003 month: '05' oa: 1 oa_version: Submitted Version page: 991 - 999 publication_status: published publisher: ML Research Press publist_id: '4844' quality_controlled: '1' scopus_import: '1' status: public title: A PAC-Bayesian bound for Lifelong Learning type: conference user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87 volume: 32 year: '2014' ... --- _id: '2294' abstract: - lang: eng text: "In this work we propose a system for automatic classification of Drosophila embryos into developmental stages.\r\nWhile the system is designed to solve an actual problem in biological research, we believe that the principle underly-\r\ning it is interesting not only for biologists, but also for researchers in computer vision. The main idea is to combine two orthogonal sources of information: one is a classifier trained on strongly invariant features, which makes it applicable to images of very different conditions, but also leads to rather noisy predictions. The other is a label propagation step based on a more powerful similarity measure that however is only consistent within specific subsets of the data at a time.\r\nIn our biological setup, the information sources are the shape and the staining patterns of embryo images. We show\r\nexperimentally that while neither of the methods \ can be used by itself to achieve satisfactory results, their combina-\r\ntion achieves prediction quality comparable to human performance." author: - first_name: Tomas full_name: Kazmar, Tomas last_name: Kazmar - first_name: Evgeny full_name: Kvon, Evgeny last_name: Kvon - first_name: Alexander full_name: Stark, Alexander last_name: Stark - first_name: Christoph full_name: Lampert, Christoph id: 40C20FD2-F248-11E8-B48F-1D18A9856A87 last_name: Lampert orcid: 0000-0001-8622-7887 citation: ama: 'Kazmar T, Kvon E, Stark A, Lampert C. Drosophila Embryo Stage Annotation using Label Propagation. In: IEEE; 2013. doi:10.1109/ICCV.2013.139' apa: 'Kazmar, T., Kvon, E., Stark, A., & Lampert, C. (2013). Drosophila Embryo Stage Annotation using Label Propagation. Presented at the ICCV: International Conference on Computer Vision, Sydney, Australia: IEEE. https://doi.org/10.1109/ICCV.2013.139' chicago: Kazmar, Tomas, Evgeny Kvon, Alexander Stark, and Christoph Lampert. “Drosophila Embryo Stage Annotation Using Label Propagation.” IEEE, 2013. https://doi.org/10.1109/ICCV.2013.139. ieee: 'T. Kazmar, E. Kvon, A. Stark, and C. Lampert, “Drosophila Embryo Stage Annotation using Label Propagation,” presented at the ICCV: International Conference on Computer Vision, Sydney, Australia, 2013.' ista: 'Kazmar T, Kvon E, Stark A, Lampert C. 2013. Drosophila Embryo Stage Annotation using Label Propagation. ICCV: International Conference on Computer Vision.' mla: Kazmar, Tomas, et al. Drosophila Embryo Stage Annotation Using Label Propagation. IEEE, 2013, doi:10.1109/ICCV.2013.139. short: T. Kazmar, E. Kvon, A. Stark, C. Lampert, in:, IEEE, 2013. conference: end_date: 2013-12-08 location: Sydney, Australia name: 'ICCV: International Conference on Computer Vision' start_date: 2013-12-01 date_created: 2018-12-11T11:56:49Z date_published: 2013-12-01T00:00:00Z date_updated: 2021-01-12T06:56:35Z day: '01' department: - _id: ChLa doi: 10.1109/ICCV.2013.139 ec_funded: 1 language: - iso: eng main_file_link: - open_access: '1' url: http://www.cv-foundation.org/openaccess/ICCV2013.py month: '12' oa: 1 oa_version: Submitted Version project: - _id: 2532554C-B435-11E9-9278-68D0E5697425 call_identifier: FP7 grant_number: '308036' name: Lifelong Learning of Visual Scene Understanding publication_status: published publisher: IEEE publist_id: '4634' quality_controlled: '1' scopus_import: 1 status: public title: Drosophila Embryo Stage Annotation using Label Propagation type: conference user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87 year: '2013' ... --- _id: '2293' abstract: - lang: eng text: Many computer vision problems have an asymmetric distribution of information between training and test time. In this work, we study the case where we are given additional information about the training data, which however will not be available at test time. This situation is called learning using privileged information (LUPI). We introduce two maximum-margin techniques that are able to make use of this additional source of information, and we show that the framework is applicable to several scenarios that have been studied in computer vision before. Experiments with attributes, bounding boxes, image tags and rationales as additional information in object classification show promising results. author: - first_name: Viktoriia full_name: Sharmanska, Viktoriia id: 2EA6D09E-F248-11E8-B48F-1D18A9856A87 last_name: Sharmanska orcid: 0000-0003-0192-9308 - first_name: Novi full_name: Quadrianto, Novi last_name: Quadrianto - first_name: Christoph full_name: Lampert, Christoph id: 40C20FD2-F248-11E8-B48F-1D18A9856A87 last_name: Lampert orcid: 0000-0001-8622-7887 citation: ama: 'Sharmanska V, Quadrianto N, Lampert C. Learning to rank using privileged information. In: IEEE; 2013:825-832. doi:10.1109/ICCV.2013.107' apa: 'Sharmanska, V., Quadrianto, N., & Lampert, C. (2013). Learning to rank using privileged information (pp. 825–832). Presented at the ICCV: International Conference on Computer Vision, Sydney, Australia: IEEE. https://doi.org/10.1109/ICCV.2013.107' chicago: Sharmanska, Viktoriia, Novi Quadrianto, and Christoph Lampert. “Learning to Rank Using Privileged Information,” 825–32. IEEE, 2013. https://doi.org/10.1109/ICCV.2013.107. ieee: 'V. Sharmanska, N. Quadrianto, and C. Lampert, “Learning to rank using privileged information,” presented at the ICCV: International Conference on Computer Vision, Sydney, Australia, 2013, pp. 825–832.' ista: 'Sharmanska V, Quadrianto N, Lampert C. 2013. Learning to rank using privileged information. ICCV: International Conference on Computer Vision, 825–832.' mla: Sharmanska, Viktoriia, et al. Learning to Rank Using Privileged Information. IEEE, 2013, pp. 825–32, doi:10.1109/ICCV.2013.107. short: V. Sharmanska, N. Quadrianto, C. Lampert, in:, IEEE, 2013, pp. 825–832. conference: end_date: 2013-12-08 location: Sydney, Australia name: 'ICCV: International Conference on Computer Vision' start_date: 2013-12-01 date_created: 2018-12-11T11:56:49Z date_published: 2013-12-01T00:00:00Z date_updated: 2023-02-23T10:36:41Z day: '01' department: - _id: ChLa doi: 10.1109/ICCV.2013.107 ec_funded: 1 language: - iso: eng main_file_link: - open_access: '1' url: www.cv-foundation.org/openaccess/content_iccv_2013/papers/Sharmanska_Learning_to_Rank_2013_ICCV_paper.pdf month: '12' oa: 1 oa_version: Submitted Version page: 825 - 832 project: - _id: 2532554C-B435-11E9-9278-68D0E5697425 call_identifier: FP7 grant_number: '308036' name: Lifelong Learning of Visual Scene Understanding publication_status: published publisher: IEEE publist_id: '4635' quality_controlled: '1' scopus_import: 1 status: public title: Learning to rank using privileged information type: conference user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87 year: '2013' ... --- _id: '2516' abstract: - lang: eng text: 'We study the problem of object recognition for categories for which we have no training examples, a task also called zero-data or zero-shot learning. This situation has hardly been studied in computer vision research, even though it occurs frequently: the world contains tens of thousands of different object classes and for only few of them image collections have been formed and suitably annotated. To tackle the problem we introduce attribute-based classification: objects are identified based on a high-level description that is phrased in terms of semantic attributes, such as the object''s color or shape. Because the identification of each such property transcends the specific learning task at hand, the attribute classifiers can be pre-learned independently, e.g. from existing image datasets unrelated to the current task. Afterwards, new classes can be detected based on their attribute representation, without the need for a new training phase. In this paper we also introduce a new dataset, Animals with Attributes, of over 30,000 images of 50 animal classes, annotated with 85 semantic attributes. Extensive experiments on this and two more datasets show that attribute-based classification indeed is able to categorize images without access to any training images of the target classes.' author: - first_name: Christoph full_name: Lampert, Christoph id: 40C20FD2-F248-11E8-B48F-1D18A9856A87 last_name: Lampert orcid: 0000-0001-8622-7887 - first_name: Hannes full_name: Nickisch, Hannes last_name: Nickisch - first_name: Stefan full_name: Harmeling, Stefan last_name: Harmeling citation: ama: Lampert C, Nickisch H, Harmeling S. Attribute-based classification for zero-shot learning of object categories. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2013;36(3):453-465. doi:10.1109/TPAMI.2013.140 apa: Lampert, C., Nickisch, H., & Harmeling, S. (2013). Attribute-based classification for zero-shot learning of object categories. IEEE Transactions on Pattern Analysis and Machine Intelligence. IEEE. https://doi.org/10.1109/TPAMI.2013.140 chicago: Lampert, Christoph, Hannes Nickisch, and Stefan Harmeling. “Attribute-Based Classification for Zero-Shot Learning of Object Categories.” IEEE Transactions on Pattern Analysis and Machine Intelligence. IEEE, 2013. https://doi.org/10.1109/TPAMI.2013.140. ieee: C. Lampert, H. Nickisch, and S. Harmeling, “Attribute-based classification for zero-shot learning of object categories,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 36, no. 3. IEEE, pp. 453–465, 2013. ista: Lampert C, Nickisch H, Harmeling S. 2013. Attribute-based classification for zero-shot learning of object categories. IEEE Transactions on Pattern Analysis and Machine Intelligence. 36(3), 453–465. mla: Lampert, Christoph, et al. “Attribute-Based Classification for Zero-Shot Learning of Object Categories.” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 36, no. 3, IEEE, 2013, pp. 453–65, doi:10.1109/TPAMI.2013.140. short: C. Lampert, H. Nickisch, S. Harmeling, IEEE Transactions on Pattern Analysis and Machine Intelligence 36 (2013) 453–465. date_created: 2018-12-11T11:58:08Z date_published: 2013-07-30T00:00:00Z date_updated: 2021-01-12T06:57:58Z day: '30' department: - _id: ChLa doi: 10.1109/TPAMI.2013.140 intvolume: ' 36' issue: '3' language: - iso: eng month: '07' oa_version: None page: 453 - 465 publication: IEEE Transactions on Pattern Analysis and Machine Intelligence publication_status: published publisher: IEEE publist_id: '4385' quality_controlled: '1' scopus_import: 1 status: public title: Attribute-based classification for zero-shot learning of object categories type: journal_article user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87 volume: 36 year: '2013' ...