# CANZSL: Cycle-Consistent Adversarial Networks for Zero-Shot Learning from Natural Language

arXiv: https://arxiv.org/abs/1909.09822 · HTML: https://mlsanity.com/paper/1909.09822/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Sep 2019.

## Built on by (most cited first, 23 of 23)

- [Breaking the Dilemma of Medical Image-to-image Translation](https://mlsanity.com/paper/2110.06465/) (2021)
- [Semantics Disentangling for Generalized Zero-Shot Learning](https://mlsanity.com/paper/2101.07978/) (2021)
- [Zero-Shot Learning by Harnessing Adversarial Samples](https://mlsanity.com/paper/2308.00313/) (2023)
- [Large-Scale Zero-Shot Image Classification from Rich and Diverse Textual Descriptions](https://mlsanity.com/paper/2103.09669/) (2021)
- [Mitigating Generation Shifts for Generalized Zero-Shot Learning](https://mlsanity.com/paper/2107.03163/) (2021)
- [GSMFlow: Generation Shifts Mitigating Flow for Generalized Zero-Shot Learning](https://mlsanity.com/paper/2207.01798/) (2022)
- [Continual Text-to-Video Retrieval with Frame Fusion and Task-Aware Routing](https://mlsanity.com/paper/2503.10111/) (2025)
- [Entropy-Based Uncertainty Calibration for Generalized Zero-Shot Learning](https://mlsanity.com/paper/2101.03292/) (2021)
- [Cycle Consistency Driven Object Discovery](https://mlsanity.com/paper/2306.02204/) (2023)
- [PEEB: Part-based Image Classifiers with an Explainable and Editable Language Bottleneck](https://mlsanity.com/paper/2403.05297/) (2024)
- [Cluster-Aware Prompt Ensemble Learning for Few-Shot Vision-Language Model Adaptation](https://mlsanity.com/paper/2510.09867/) (2025)
- [Region Semantically Aligned Network for Zero-Shot Learning](https://mlsanity.com/paper/2110.07130/) (2021)
- [SVIP: Semantically Contextualized Visual Patches for Zero-Shot Learning](https://mlsanity.com/paper/2503.10252/) (2025)
- [Cross Knowledge-based Generative Zero-Shot Learning Approach with Taxonomy Regularization](https://mlsanity.com/paper/2101.09892/) (2021)
- [On the Element-Wise Representation and Reasoning in Zero-Shot Image Recognition: A Systematic Survey](https://mlsanity.com/paper/2408.04879/) (2024)
- [Distributed Zero-Shot Learning for Visual Recognition](https://mlsanity.com/paper/2511.08170/) (2025)
- [Accelerating Multimodal Large Language Models with Prior-Corrected Token Reduction](https://mlsanity.com/paper/2606.24156/) (2026)
- [StyleRec: A Benchmark Dataset for Prompt Recovery in Writing Style Transformation](https://mlsanity.com/paper/2504.04373/) (2025)
- [Spectral Evolution-Guided Token Pruning in Multimodal Large Language Models](https://mlsanity.com/paper/2606.24165/) (2026)
- [Multi-Knowledge Fusion for New Feature Generation in Generalized Zero-Shot Learning](https://mlsanity.com/paper/2102.11566/) (2021)
- [Class Knowledge Overlay to Visual Feature Learning for Zero-Shot Image Classification](https://mlsanity.com/paper/2102.13322/) (2021)
- [Fine-Grained Zero-Shot Learning with Attribute-Centric Representations](https://mlsanity.com/paper/2512.12219/) (2025)
- [StomataSeg: Semi-Supervised Instance Segmentation for Sorghum Stomatal Components](https://mlsanity.com/paper/2602.00703/) (2026)

## Abstract (arXiv)

Existing methods using generative adversarial approaches for Zero-Shot Learning (ZSL) aim to generate realistic visual features from class semantics by a single generative network, which is highly under-constrained. As a result, the previous methods cannot guarantee that the generated visual features can truthfully reflect the corresponding semantics. To address this issue, we propose a novel method named Cycle-consistent Adversarial Networks for Zero-Shot Learning (CANZSL). It encourages a visual feature generator to synthesize realistic visual features from semantics, and then inversely translate back synthesized the visual feature to corresponding semantic space by a semantic feature generator. Furthermore, in this paper a more challenging and practical ZSL problem is considered where the original semantics are from natural language with irrelevant words instead of clean semantics that are widely used in previous work. Specifically, a multi-modal consistent bidirectional generative adversarial network is trained to handle unseen instances by leveraging noise in the natural language. A forward one-to-many mapping from one text description to multiple visual features is coupled with an inverse many-to-one mapping from the visual space to the semantic space. Thus, a multi-modal cycle-consistency loss between the synthesized semantic representations and the ground truth can be learned and leveraged to enforce the generated semantic features to approximate to the real distribution in semantic space. Extensive experiments are conducted to demonstrate that our method consistently outperforms state-of-the-art approaches on natural language-based zero-shot learning tasks.