# Improved Feature Generating Framework for Transductive Zero-shot Learning

arXiv: https://arxiv.org/abs/2412.18282 · HTML: https://mlsanity.com/paper/2412.18282/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 44 references that are in the map.
- First posted on arXiv 24 Dec 2024.

## Builds on (most cited first, 11 of 11)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Improved Training of Wasserstein GANs](https://mlsanity.com/paper/1704.00028/) (2017)
- [Decoupled Weight Decay Regularization](https://mlsanity.com/paper/1711.05101/) (2017)
- [Detecting and Correcting for Label Shift with Black Box Predictors](https://mlsanity.com/paper/1802.03916/) (2018)
- [Attribute Prototype Network for Zero-Shot Learning](https://mlsanity.com/paper/2008.08290/) (2020)
- [HSVA: Hierarchical Semantic-Visual Adaptation for Zero-Shot Learning](https://mlsanity.com/paper/2109.15163/) (2021)
- [Improving Zero-Shot Generalization for CLIP with Synthesized Prompts](https://mlsanity.com/paper/2307.07397/) (2023)
- [Transductive Zero-Shot Learning with Visual Structure Constraint](https://mlsanity.com/paper/1901.01570/) (2019)
- [I2DFormer: Learning Image to Document Attention for Zero-Shot Image Classification](https://mlsanity.com/paper/2209.10304/) (2022)
- [ZeroDiff: Solidified Visual-Semantic Correlation in Zero-Shot Learning](https://mlsanity.com/paper/2406.02929/) (2024)

## Abstract (arXiv)

Feature Generative Adversarial Networks have emerged as powerful generative models in producing high-quality representations of unseen classes within the scope of Zero-shot Learning (ZSL). This paper delves into the pivotal influence of unseen class priors within the framework of transductive ZSL (TZSL) and illuminates the finding that even a marginal prior bias can result in substantial accuracy declines. Our extensive analysis uncovers that this inefficacy fundamentally stems from the utilization of an unconditional unseen discriminator - a core component in existing TZSL. We further establish that the detrimental effects of this component are inevitable unless the generator perfectly fits class-specific distributions. Building on these insights, we introduce our Improved Feature Generation Framework, termed I-VAEGAN, which incorporates two novel components: Pseudo-conditional Feature Adversarial (PFA) learning and Variational Embedding Regression (VER). PFA circumvents the need for prior estimation by explicitly injecting the predicted semantics as pseudo conditions for unseen classes premised by precise semantic regression. Meanwhile, VER utilizes reconstructive pre-training to learn class statistics, obtaining better semantic regression. Our I-VAEGAN achieves state-of-the-art TZSL accuracy across various benchmarks and priors. Our code would be released upon acceptance.