CVLGApr 25, 2022

Zero-Shot Logit Adjustment

arXiv:2204.11822v421 citationsh-index: 117Has Code
Originality Incremental advance
AI Analysis

This work addresses a specific bottleneck in zero-shot learning for computer vision, offering an incremental improvement by enhancing classifier training in existing generative frameworks.

The paper tackles the challenge of improving classifier performance in generation-based Generalized Zero-Shot Learning by analyzing bias and homogeneity in generated pseudo-unseen samples and incorporating these as priors via logit adjustment, achieving state-of-the-art results when combined with a basic generator.

Semantic-descriptor-based Generalized Zero-Shot Learning (GZSL) poses challenges in recognizing novel classes in the test phase. The development of generative models enables current GZSL techniques to probe further into the semantic-visual link, culminating in a two-stage form that includes a generator and a classifier. However, existing generation-based methods focus on enhancing the generator's effect while neglecting the improvement of the classifier. In this paper, we first analyze of two properties of the generated pseudo unseen samples: bias and homogeneity. Then, we perform variational Bayesian inference to back-derive the evaluation metrics, which reflects the balance of the seen and unseen classes. As a consequence of our derivation, the aforementioned two properties are incorporated into the classifier training as seen-unseen priors via logit adjustment. The Zero-Shot Logit Adjustment further puts semantic-based classifiers into effect in generation-based GZSL. Our experiments demonstrate that the proposed technique achieves state-of-the-art when combined with the basic generator, and it can improve various generative Zero-Shot Learning frameworks. Our codes are available on https://github.com/cdb342/IJCAI-2022-ZLA.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes