Abstract:Zero-Shot Learning (ZSL) involves transferring knowledge from seen classes to unseen classes by establishing connections between visual and semantic spaces. Traditional ZSL methods identify novel classes by class-level attribute vectors, which implies an information bottleneck. These approaches often use class-level attribute vectors as the fitting target during training, disregarding the individual variations within a class. Moreover, the attributes used for training lack location information and are prone to mismatch with local regions of visual features. To this end, we introduce a Near-Instance-Level Attribute Bottleneck (IAB) to alter class-level attribute vectors as well as visual features throughout the training phase to better reflect their naturalistic correspondences. Specifically, our Near-Instance-Wise Attribute Adaptation (NAA) modifies class attribute vectors to obtain multiple attribute basis vectors, generating a subspace that is more relevant to instance-level samples. Additionally, our Vision Attribute Relation Strengthening (VARS) module searches for attribute-related regions within the features, offering additional location information during the training phase. The proposed method is evaluated on four ZSL benchmarks, revealing that it is superior or competitive to the state-of-the-art methods on ZSL and the more challenging Generalized Zero-Shot Learning (GZSL) settings. Extensive experiments corroborate the sustainability of this study as one of the most potential directions for ZSL, i.e. , the effectiveness of enhancing the visual-semantic relationships formed during training using a simple model structure. Code is available at: https://github.com/LanchJL/IAB-GZSL.

Complementary Attributes: A New Clue to Zero-Shot Learning

Learning complementary semantic information for zero-shot recognition

Learning Conditional Attributes for Compositional Zero-Shot Learning

Boosting Zero-shot Learning via Contrastive Optimization of Attribute Representations

COMAE: COMprehensive Attribute Exploration for Zero-shot Hashing

Attribute self-representation steered by exclusive lasso for zero-shot learning

OntoZSL: Ontology-enhanced Zero-shot Learning

Selective Zero-Shot Classification with Augmented Attributes.

Learning object-centric complementary features for zero-shot learning

MAC: A Benchmark for Multiple Attributes Compositional Zero-Shot Learning

Improving Generalization via Attribute Selection on Out-of-the-box Data

Learning Discriminative Latent Attributes for Zero-Shot Classification.

Compositional Zero-Shot Learning with Contextualized Cues and Adaptive Contrastive Training

Zero-Shot Learning with Attribute Selection.

Zero-shot Learning with a Partial Set of Observed Attributes

Prior Knowledge about Attributes: Learning a More Effective Potential Space for Zero-Shot Recognition

Hybrid Discriminative Attribute-Object Embedding Network for Compositional Zero-Shot Learning

Estimation of Near-Instance-Level Attribute Bottleneck for Zero-Shot Learning

Attribute subspaces for zero-shot learning

A Robust Generalized Zero-Shot Learning Method with Attribute Prototype and Discriminative Attention Mechanism

Visual-guided attentive attributes embedding for zero-shot learning