mlsanityData to 13 Oct 2026

Revisiting Document Representations for Large-Scale Zero-Shot Learning

arXiv 2104.10355 ↗First posted 21 Apr 2021cs.CV cs.AI cs.CLcited by 9

Not on the map. Why?

Abstract (arXiv)

Zero-shot learning aims to recognize unseen objects using their semantic representations. Most existing works use visual attributes labeled by humans, not suitable for large-scale applications. In this paper, we revisit the use of documents as semantic representations. We argue that documents like Wikipedia pages contain rich visual information, which however can easily be buried by the vast amount of non-visual sentences. To address this issue, we propose a semi-automatic mechanism for visual sentence extraction that leverages the document section headers and the clustering structure of visual sentences. The extracted visual sentences, after a novel weighting scheme to distinguish similar classes, essentially form semantic representations like visual attributes but need much less human effort. On the ImageNet dataset with over 10,000 unseen classes, our representations lead to a 64% relative improvement against the commonly used ones.

Key facts. Cited by 9 papers in the map (data through 13 Oct 2026). Builds on 30 of its 46 references that are in the map. First posted on arXiv 21 Apr 2021.

Builds on

30 of 46 references.

References · 30

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,950
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,304
  3. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,107
  4. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,476
  5. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,213
  6. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,467
  7. 2016Generative Adversarial Text to Image Synthesiscited by 735
  8. 2017Zero-Shot Learning -- A Comprehensive Evaluation of the Good, the Bad and the Uglycited by 538
  9. 2019Revealing the Dark Secrets of BERTcited by 300
  10. 2017Feature Generating Networks for Zero-Shot Learningcited by 257
  11. 2013Zero-Shot Learning Through Cross-Modal Transfercited by 237
  12. 2018Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoderscited by 206
18 more
  1. 2016Learning Deep Representations of Fine-grained Visual Descriptionscited by 182
  2. 2014Evaluation of Output Embeddings for Fine-Grained Image Classificationcited by 166
  3. 2015Label-Embedding for Image Classificationcited by 164
  4. 2019f-VAEGAN-D2: A Feature Generating Framework for Any-Shot Learningcited by 164
  5. 2018Zero-shot Recognition via Semantic Embeddings and Knowledge Graphscited by 155
  6. 2013Zero-Shot Learning by Convex Combination of Semantic Embeddingscited by 135
  7. 2017Semantic Autoencoder for Zero-Shot Learningcited by 133
  8. 2017A Generative Adversarial Approach for Zero-Shot Learning from Noisy Textscited by 96
  9. 2018Rethinking Knowledge Graph Propagation for Zero-Shot Learningcited by 87
  10. 2016Less is more: zero-shot learning from online textual documents with noise suppressioncited by 40
  11. 2016Predicting Visual Exemplars of Unseen Classes for Zero-Shot Learningcited by 30
  12. 2018A Large-scale Attribute Dataset for Zero-shot Learningcited by 21
  13. 2018Generalized Zero-Shot Recognition based on Visually Semantic Embeddingcited by 19
  14. 2018Quantifying the visual concreteness of words and topics in multimodal datasetscited by 16
  15. 2017Automatic Discovery, Association Estimation and Learning of Semantic Attributes for a Thousand Categoriescited by 10
  16. 2018Classifier and Exemplar Synthesis for Zero-Shot Learningcited by 9
  17. 2015Unsupervised Learning on Neural Network Outputs: with Application in Zero-shot Learningcited by 4
  18. 2020Webly Supervised Semantic Embeddings for Large Scale Zero-Shot Learningcited by 3

Built on by

Cited by year

Citing papers per year: 2022: 4; 2023: 2; 2024: 1; 2025: 2.

9 papers.

  1. 2022Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classificationcited by 311
  2. 2023Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entitiescited by 112
  3. 2022Decoding Visual Neural Representations by Multimodal Learning of Brain-Visual-Linguistic Featurescited by 63
  4. 2023DocILE Benchmark for Document Information Localization and Extractioncited by 58
  5. 2022I2MVFormer: Large Language Model Generated Multi-View Document Supervision for Zero-Shot Image Classificationcited by 54
  6. 2022I2DFormer: Learning Image to Document Attention for Zero-Shot Image Classificationcited by 27
  7. 2024Visual-Semantic Decomposition and Partial Alignment for Document-based Zero-Shot Learningcited by 4
  8. 2025Coreset Selection via LLM-based Concept Bottleneckscited by 2
  9. 2025MADS: Multi-Attribute Document Supervision for Zero-Shot Image Classificationcited by 2