# Towards Effective Deep Embedding for Zero-Shot Learning

arXiv: https://arxiv.org/abs/1808.10075 · HTML: https://mlsanity.com/paper/1808.10075/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 30 Aug 2018.

## Built on by (most cited first, 19 of 19)

- [A Closer Look at Self-training for Zero-Label Semantic Segmentation](https://mlsanity.com/paper/2104.11692/) (2021)
- [Zero-Shot Learning on 3D Point Cloud Objects and Beyond](https://mlsanity.com/paper/2104.04980/) (2021)
- [Multi-Semantic Fusion Model for Generalized Zero-Shot Skeleton-Based Action Recognition](https://mlsanity.com/paper/2309.09592/) (2023)
- [Region Semantically Aligned Network for Zero-Shot Learning](https://mlsanity.com/paper/2110.07130/) (2021)
- [Sentinel-Guided Zero-Shot Learning: A Collaborative Paradigm without Real Data Exposure](https://mlsanity.com/paper/2403.09363/) (2024)
- [An Iterative Co-Training Transductive Framework for Zero Shot Learning](https://mlsanity.com/paper/2203.16041/) (2022)
- [Bridging the Projection Gap: Overcoming Projection Bias Through Parameterized Distance Learning](https://mlsanity.com/paper/2309.01390/) (2023)
- [Absolute Zero-Shot Learning](https://mlsanity.com/paper/2202.11319/) (2022)
- [LETS-GZSL: A Latent Embedding Model for Time Series Generalized Zero Shot Learning](https://mlsanity.com/paper/2207.12007/) (2022)
- [Transductive Zero-Shot Learning by Decoupled Feature Generation](https://mlsanity.com/paper/2102.03266/) (2021)
- [Semantic-diversity transfer network for generalized zero-shot learning via inner disagreement based OOD detector](https://mlsanity.com/paper/2203.09017/) (2022)
- [Learning Compositional Representations for Effective Low-Shot Generalization](https://mlsanity.com/paper/2204.08090/) (2022)
- [Fuzzy Logic Visual Network (FLVN): A neuro-symbolic approach for visual features matching](https://mlsanity.com/paper/2307.16019/) (2023)
- [Semantic Borrowing for Generalized Zero-Shot Learning](https://mlsanity.com/paper/2102.04969/) (2021)
- [A Simple Approach for Zero-Shot Learning based on Triplet Distribution Embeddings](https://mlsanity.com/paper/2103.15939/) (2021)
- [Zero-sample surface defect detection and classification based on semantic feedback neural network](https://mlsanity.com/paper/2106.07959/) (2021)
- [Prototypical Model with Novel Information-theoretic Loss Function for Generalized Zero Shot Learning](https://mlsanity.com/paper/2112.03134/) (2021)
- [On the Transferability of Visual Features in Generalized Zero-Shot Learning](https://mlsanity.com/paper/2211.12494/) (2022)
- [CAST: Closed-form Analytic Semantic Transfer for Zero-Shot Classifier Extension](https://mlsanity.com/paper/2608.13751/) (2026)

## Abstract (arXiv)

Zero-shot learning (ZSL) can be formulated as a cross-domain matching problem: after being projected into a joint embedding space, a visual sample will match against all candidate class-level semantic descriptions and be assigned to the nearest class. In this process, the embedding space underpins the success of such matching and is crucial for ZSL. In this paper, we conduct an in-depth study on the construction of embedding space for ZSL and posit that an ideal embedding space should satisfy two criteria: intra-class compactness and inter-class separability. While the former encourages the embeddings of visual samples of one class to distribute tightly close to the semantic description embedding of this class, the latter requires embeddings from different classes to be well separated from each other. Towards this goal, we present a simple but effective two-branch network to simultaneously map semantic descriptions and visual samples into a joint space, on which visual embeddings are forced to regress to their class-level semantic embeddings and the embeddings crossing classes are required to be distinguishable by a trainable classifier. Furthermore, we extend our method to a transductive setting to better handle the model bias problem in ZSL (i.e., samples from unseen classes tend to be categorized into seen classes) with minimal extra supervision. Specifically, we propose a pseudo labeling strategy to progressively incorporate the testing samples into the training process and thus balance the model between seen and unseen classes. Experimental results on five standard ZSL datasets show the superior performance of the proposed method and its transductive extension.