Learning by Imagining: Debiased Feature Augmentation for Compositional Zero-Shot Learning
Compositional Zero-Shot Learning (CZSL) aims to recognize unseen attribute-object compositions by learning prior knowledge of seen primitives, \textit{i.e.}, attributes and objects. Learning generalizable compositional representations in CZSL remains challenging due to the entangled nature of attributes and objects as well as the prevalence of long-tailed distributions in real-world data. Inspired by neuroscientific findings that imagination and perception share similar neural processes, we propose a novel approach called Debiased Feature Augmentation (DeFA) to address these challenges. The proposed DeFA integrates a disentangle-and-reconstruct framework for feature augmentation with a debiasing strategy. DeFA explicitly leverages the prior knowledge of seen attributes and objects by synthesizing high-fidelity composition features to support compositional generalization. Extensive experiments on three widely used datasets demonstrate that DeFA achieves state-of-the-art performance in both \textit{closed-world} and \textit{open-world} settings.
Code (0)
등록된 구현이 없습니다.
Tasks
Compositional Zero-Shot LearningSimilar Papers 제목 키워드 기반
Bias-Eliminating Augmentation Learning for Debiased Federated Learning
Learning models trained on biased datasets tend to observe correlations between categorical and undesirable features, which result in degraded performances. Most existing debiased learning models are designed for cen…
Federated Learningimage-classificationImage ClassificationCompositional Zero-Shot Domain Transfer with Text-to-Text Models
Label scarcity is a bottleneck for improving task performance in specialised domains. We propose a novel compositional transfer learning framework (DoT5 - domain compositional zero-shot T5) for zero-shot domain transfer.…
Data AugmentationMulti-Task LearningTransfer LearningDiversified Augmentation with Domain Adaptation for Debiased Video Temporal Grounding
Temporal sentence grounding in videos (TSGV) faces challenges due to public TSGV datasets containing significant temporal biases, which are attributed to the uneven temporal distributions of target moments. Existing meth…
Data AugmentationDomain AdaptationSentenceTemporal Sentence GroundingImagining an Engineer: On GAN-Based Data Augmentation Perpetuating Biases
The use of synthetic data generated by Generative Adversarial Networks (GANs) has become quite a popular method to do data augmentation for many applications. While practitioners celebrate this as an economical way to ge…
Data AugmentationTask-Aware Feature Generation for Zero-Shot Compositional Learning
Visual concepts (e.g., red apple, big elephant) are often semantically compositional and each element of the compositions can be reused to construct novel concepts (e.g., red elephant). Compositional feature synthesis, w…
Novel ConceptsZero-Shot Learning