BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//EE - ECPv5.10.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:EE
X-ORIGINAL-URL:https://ee.iisc.ac.in
X-WR-CALDESC:Events for EE
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
BEGIN:STANDARD
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
TZNAME:IST
DTSTART:20220101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Kolkata:20220926T163000
DTEND;TZID=Asia/Kolkata:20220926T173000
DTSTAMP:20260421T135032
CREATED:20220926T010719Z
LAST-MODIFIED:20220926T010758Z
UID:240014-1664209800-1664213400@ee.iisc.ac.in
SUMMARY:Thesis Defence of Mr. Ahmad Arfeen @ 11am
DESCRIPTION:Title:  Data Efficient Domain Generalization \nFaculty Advisor: Prof. Soma Biswas \nExaminer: Prof. Venkatesh Babu R. (CDS\, IISc) \nDate: Monday\, September 26\, 2022 \nTime: 11:00 AM \nVenue:  MMCR (EE dept) \nAbstract: For the task of image classification\, in general\, the test data is assumed to come from the same distribution as the training data. But this may not always hold in real-life scenarios. For example\, in night-time surveillance\, we may need to classify images captured using NIR cameras\, but the available model has been trained on RGB images. Domain generalization (DG) addresses the problem of generalizing classification performance across any unknown domain\, by leveraging training samples from multiple source domains. In this thesis\, we address two challenging scenarios for the DG task\, with focus on data efficiency. Currently\, the training process of majority of the state-of-the-art DG-methods is dependent on a large amount of labeled data. This restricts the application of the models in many real-world scenarios\, where collecting and annotating a large dataset is an expensive and difficult task. \nAs the first contribution\, we address the problem of Semi-supervised Domain Generalization (SSDG)\, where the training set contains only a few labeled data\, in addition to a large number of unlabeled data from multiple domains. This is relatively unexplored in literature and poses a considerable challenge to the state-of-the-art DG models\, since their performance degrades under such condition. To address this scenario\, we propose a novel Selective Mixing and Voting Network (SMV-Net)\, which effectively extracts useful knowledge from the set of unlabeled training data\, available to the model. Specifically\, we propose a mixing strategy on selected unlabeled samples on which the model is confident about their predicted class labels to achieve a domain-invariant representation of the data\, which generalizes effectively across any unseen domain. Extensive experiments on two popular DG-datasets demonstrate the usefulness of the proposed framework. \nThe second contribution of this thesis is a novel approach for the task of Zero-Shot Domain Generalization (ZSDG). This is very challenging since the query data can belong to an unseen class as well as unseen domain. For this task\, we address the challenge of class imbalance by learning class specific classifier margins\, which not only maintains the semantic relationship of the classes in the embedding space\, but is also discriminative\, and thus improves the classification performance on the test data. Extensive experiments on multiple datasets justify the effectiveness of the proposed approach. \n*************************************************************************************************** \nALL ARE WELCOME
URL:https://ee.iisc.ac.in/event/thesis-defence-of-mr-%e2%80%afahmad-arfeen-11am/
LOCATION:EE\, MMCR
END:VEVENT
END:VCALENDAR