mlsanityData to 7 Oct 2026

Relation-aware Compositional Zero-shot Learning for Attribute-Object Pair Recognition

arXiv 2108.04603 ↗First posted 10 Aug 2021cs.CV cs.MMcited by 19

Not on the map. Why?

Abstract (arXiv)

This paper proposes a novel model for recognizing images with composite attribute-object concepts, notably for composite concepts that are unseen during model training. We aim to explore the three key properties required by the task --- relation-aware, consistent, and decoupled --- to learn rich and robust features for primitive concepts that compose attribute-object pairs. To this end, we propose the Blocked Message Passing Network (BMP-Net). The model consists of two modules. The concept module generates semantically meaningful features for primitive concepts, whereas the visual module extracts visual features for attributes and objects from input images. A message passing mechanism is used in the concept module to capture the relations between primitive concepts. Furthermore, to prevent the model from being biased towards seen composite concepts and reduce the entanglement between attributes and objects, we propose a blocking mechanism that equalizes the information available to the model for both seen and unseen concepts. Extensive experiments and ablation studies on two benchmarks show the efficacy of the proposed model.

Key facts. Cited by 19 papers in the map (data through 7 Oct 2026). Builds on 42 of its 87 references that are in the map. First posted on arXiv 10 Aug 2021.

Builds on

42 of 87 references.

References · 42

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  4. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,530
  5. 2013Auto-Encoding Variational Bayescited by 3,980
  6. 2019ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Taskscited by 2,219
  7. 2016Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoisingcited by 1,352
  8. 2017Zero-Shot Learning -- A Comprehensive Evaluation of the Good, the Bad and the Uglycited by 536
  9. 2015Towards Open Set Deep Networkscited by 533
  10. 201912-in-1: Multi-Task Vision and Language Representation Learningcited by 278
  11. 2017Feature Generating Networks for Zero-Shot Learningcited by 257
  12. 2016Key-Value Memory Networks for Directly Reading Documentscited by 223
30 more
  1. 2016FractalNet: Ultra-Deep Neural Networks without Residualscited by 216
  2. 2016A Decomposable Attention Model for Natural Language Inferencecited by 214
  3. 2018Learning Human-Object Interactions by Graph Parsing Neural Networkscited by 176
  4. 2015Inverting Visual Representations with Convolutional Networkscited by 168
  5. 2019f-VAEGAN-D2: A Feature Generating Framework for Any-Shot Learningcited by 164
  6. 2018Zero-shot Recognition via Semantic Embeddings and Knowledge Graphscited by 155
  7. 2016An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wildcited by 146
  8. 2017Semantic Autoencoder for Zero-Shot Learningcited by 133
  9. 2019Task-Driven Modular Networks for Zero-Shot Compositional Learningcited by 121
  10. 2016Synthesized Classifiers for Zero-Shot Learningcited by 113
  11. 2020Symmetry and Group in Attribute-Object Compositionscited by 89
  12. 2020Zero-Shot Video Object Segmentation via Attentive Graph Neural Networkscited by 84
  13. 2020A causal view of compositional zero-shot recognitioncited by 78
  14. 2017Zero-Shot Visual Recognition using Semantics-Preserving Adversarial Embedding Networkscited by 66
  15. 2019Embedding Symbolic Knowledge into Deep Networkscited by 49
  16. 2020Invertible Zero-Shot Recognition Flowscited by 48
  17. 2019Spatiotemporal Co-attention Recurrent Neural Networks for Human-Skeleton Motion Predictioncited by 46
  18. 2016Learning Attributes Equals Multi-Source Domain Generalizationcited by 44
  19. 2019Explainable Video Action Reasoning via Prior Knowledge and State Transitionscited by 33
  20. 2020A Novel Perspective to Zero-shot Learning: Towards an Alignment of Manifold Structures via Semantic Feature Expansioncited by 31
  21. 2018Interact as You Intend: Intention-Driven Human-Object Interaction Detectioncited by 30
  22. 2017Joint Discovery of Object States and Manipulation Actionscited by 28
  23. 2017A Unified approach for Conventional Zero-shot, Generalized Zero-shot and Few-shot Learningcited by 28
  24. 2019Reasoning Visual Dialogs with Structural and Partial Observationscited by 25
  25. 2016Recovering the Missing Link: Predicting Class-Attribute Associations for Unsupervised Zero-Shot Learningcited by 20
  26. 2018Attributes as Operators: Factorizing Unseen Attribute-Object Compositionscited by 20
  27. 2018Neural Algebra of Classifierscited by 7
  28. 2019Domain-Specific Embedding Network for Zero-Shot Recognitioncited by 5
  29. 2017Joint Intermodal and Intramodal Label Transfers for Extremely Rare or Unseen Classescited by 4
  30. 2020GradMix: Multi-source Transfer across Domains and Taskscited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 5; 2023: 5; 2024: 6; 2025: 2; 2026: 1.

19 papers.

Sort:

By area:

  1. 2022Disentangling Visual Embeddings for Attributes and Objectscited by 55
  2. 2023Learning Conditional Attributes for Compositional Zero-Shot Learningcited by 28
  3. 2024Context-based and Diversity-driven Specificity in Compositional Zero-Shot Learningcited by 20
  4. 2023Hierarchical Visual Primitive Experts for Compositional Zero-Shot Learningcited by 16
  5. 2023Chop & Learn: Recognizing and Generating Object-State Compositionscited by 13
  6. 2023Distilled Reverse Attention Network for Open-world Compositional Zero-Shot Learningcited by 11
  7. 2022ProCC: Progressive Cross-primitive Compatibility for Open-World Compositional Zero-Shot Learningcited by 9
  8. 2022Distance Matters in Human-Object Interaction Detectioncited by 8
  9. 2024On the Element-Wise Representation and Reasoning in Zero-Shot Image Recognition: A Systematic Surveycited by 5
  10. 2022LOCL: Learning Object-Attribute Composition using Localizationcited by 3
  11. 2025A Conditional Probability Framework for Compositional Zero-shot Learningcited by 2
  12. 2024CSCNET: Class-Specified Cascaded Network for Compositional Zero-Shot Learningcited by 1
7 more
  1. 2024MAC: A Benchmark for Multiple Attributes Compositional Zero-Shot Learningcited by 1
  2. 2024Unified Framework for Open-World Compositional Zero-shot Learningcited by 1
  3. 2024Compositional Zero-Shot Learning with Contextualized Cues and Adaptive Contrastive Trainingcited by 1
  4. 2025Compositional Zero-Shot Learning: A Surveycited by 1
  5. 2022Simple Primitives with Feasibility- and Contextuality-Dependence for Open-World Compositional Zero-shot Learningcited by 0
  6. 2023Compositional Zero-Shot Learning for Attribute-Based Object Reference in Human-Robot Interactioncited by 0
  7. 2026Progressive Reasoning with Primitive Correction for Compositional Zero-Shot Learningcited by 0