SEIC: Semantic Embedding with Intermediate Classes for Zero-Shot Domain Generalization

Mondal, B and Biswas, S (2023) SEIC: Semantic Embedding with Intermediate Classes for Zero-Shot Domain Generalization. In: 16th Asian Conference on Computer Vision, ACCV 2022, 4-8 December 2022, Macao, pp. 333-350.

PDF accv_2022.pdf - Published Version Restricted to Registered users only Download (1MB) | Request a copy

Abstract

In this work, we address the Zero-Shot Domain Generalization (ZSDG) task, where the goal is to learn a model from multiple source domains, such that it can generalize well to both unseen classes and unseen domains during testing. Since it combines the tasks of Domain Generalization (DG) and Zero-Shot Learning (ZSL), here we explore whether advances in these fields also translate to improved performance for the ZSDG task. Specifically, we build upon a state-of-the-art approach for domain generalization and appropriately modify it such that it can generalize to unseen classes during the testing stage. Towards this goal, we propose to make the feature embedding space semantically meaningful, by not only making an image feature close to its semantic attributes, but also taking into account its similarity with the other neighbouring classes. In addition, in order to reserve space for the unseen classes in the embedding space, we propose to introduce pseudo intermediate classes in between the semantically similar classes during training. This reduces confusion of the similar classes and thus increases the discriminability of the embedding space. Extensive experiments on two large-scale benchmark datasets, namely DomainNet and DomainNet-LS and comparisons with the state-of-the-art approaches show that the proposed framework outperforms all the other techniques on both the datasets. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Item Type:	Conference Paper
Publication:	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Publisher:	Springer Science and Business Media Deutschland GmbH
Additional Information:	The copyright for this article belongs to Springer Science and Business Media Deutschland GmbH.
Keywords:	Large dataset; Semantics; Well testing; Zero-shot learning, Feature embedding; Generalisation; Image features; Learn+; Multiple source; Performance; Semantic attribute; Semantic embedding; State-of-the-art approach, Embeddings
Department/Centre:	Division of Electrical Sciences > Electrical Engineering
Date Deposited:	13 Apr 2023 10:38
Last Modified:	13 Apr 2023 10:38
URI:	https://eprints.iisc.ac.in/id/eprint/81330

Actions (login required)

View Item