mlsanityData to 7 Oct 2026

Skew Class-balanced Re-weighting for Unbiased Scene Graph Generation

arXiv 2301.00351 ↗First posted 1 Jan 2023cs.LGcited by 5

Not on the map. Why?

Abstract (arXiv)

An unbiased scene graph generation (SGG) algorithm referred to as Skew Class-balanced Re-weighting (SCR) is proposed for considering the unbiased predicate prediction caused by the long-tailed distribution. The prior works focus mainly on alleviating the deteriorating performances of the minority predicate predictions, showing drastic dropping recall scores, i.e., losing the majority predicate performances. It has not yet correctly analyzed the trade-off between majority and minority predicate performances in the limited SGG datasets. In this paper, to alleviate the issue, the Skew Class-balanced Re-weighting (SCR) loss function is considered for the unbiased SGG models. Leveraged by the skewness of biased predicate predictions, the SCR estimates the target predicate weight coefficient and then re-weights more to the biased predicates for better trading-off between the majority predicates and the minority ones. Extensive experiments conducted on the standard Visual Genome dataset and Open Image V4 & V6 show the performances and generality of the SCR with the traditional SGG models.

Key facts. Cited by 5 papers in the map (data through 7 Oct 2026). Builds on 57 of its 80 references that are in the map. First posted on arXiv 1 Jan 2023.

Builds on

57 of 80 references.

References · 57

  1. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
  2. 2016Feature Pyramid Networks for Object Detectioncited by 5,021
  3. 2016Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotationscited by 2,646
  4. 2018Graph Neural Networks: A Review of Methods and Applicationscited by 1,571
  5. 2019Class-Balanced Loss Based on Effective Number of Samplescited by 1,239
  6. 2022YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectorscited by 978
  7. 2018The Open Images Dataset V4: Unified image classification, object detection, and visual relationship detection at scalecited by 542
  8. 2017Scene Graph Generation by Iterative Message Passingcited by 462
  9. 2017Neural Motifs: Scene Graph Parsing with Global Contextcited by 413
  10. 2020Unbiased Scene Graph Generation from Biased Trainingcited by 355
  11. 2021Towards Open World Object Detectioncited by 297
  12. 2018Graph R-CNN for Scene Graph Generationcited by 296
38 more
  1. 2018Learning to Compose Dynamic Tree Structures for Visual Contextscited by 258
  2. 2021Attention Mechanisms in Computer Vision: A Surveycited by 232
  3. 2022Discovering Invariant Rationales for Graph Neural Networkscited by 186
  4. 2022GreaseLM: Graph REASoning Enhanced Language Models for Question Answeringcited by 170
  5. 2020GPS-Net: Graph Property Sensing Network for Scene Graph Generationcited by 165
  6. 2019Knowledge-Embedded Routing Network for Scene Graph Generationcited by 163
  7. 2017Scene Graph Generation from Objects, Phrases and Region Captionscited by 154
  8. 2021Bipartite Graph Network with Adaptive Message Passing for Unbiased Scene Graph Generationcited by 137
  9. 2022Panoptic Scene Graph Generationcited by 114
  10. 2021Energy-Based Learning for Scene Graph Generationcited by 104
  11. 2019Scene Graph Generation with External Knowledge and Image Reconstructioncited by 102
  12. 2019Graphical Contrastive Losses for Scene Graph Parsingcited by 100
  13. 2017Detecting Visual Relationships with Deep Relational Networkscited by 97
  14. 2022RelTR: Relation Transformer for Scene Graph Generationcited by 95
  15. 2021Do Datasets Have Politics? Disciplinary Values in Computer Vision Dataset Developmentcited by 85
  16. 2018Factorizable Net: An Efficient Subgraph-based Framework for Scene Graph Generationcited by 78
  17. 2020PCPL: Predicate-Correlation Perception Learning for Unbiased Scene Graph Generationcited by 75
  18. 2021Recovering the Unbiased Scene Graphs from the Biased Onescited by 71
  19. 2022Stacked Hybrid-Attention and Group Collaborative Learning for Unbiased Scene Graph Generationcited by 67
  20. 2021Learning of Visual Relations: The Devil is in the Tailscited by 65
  21. 2015Seeing through the Human Reporting Bias: Visual Classifiers from Noisy Human-Centric Labelscited by 64
  22. 2020CogTree: Cognition Tree Loss for Unbiased Scene Graph Generationcited by 62
  23. 2017Visual Relationship Detection with Internal and External Linguistic Knowledge Distillationcited by 60
  24. 2022The Devil is in the Labels: Noisy Label Correction for Robust Scene Graph Generationcited by 60
  25. 2021Fully Convolutional Scene Graph Generationcited by 57
  26. 2022Fine-Grained Scene Graph Generation with Data Transfercited by 57
  27. 2022Towards Open-vocabulary Scene Graph Generation with Prompt-based Finetuningcited by 54
  28. 2018Counterfactual Critic Multi-Agent Training for Scene Graph Generationcited by 53
  29. 2021From General to Specific: Informative Scene Graph Generation via Balance Adjustmentcited by 51
  30. 2018LinkNet: Relational Embedding for Scene Graphcited by 45
  31. 2021Structured Sparse R-CNN for Direct Scene Graph Generationcited by 44
  32. 2018Zoom-Net: Mining Deep Feature Interactions for Visual Relationship Recognitioncited by 41
  33. 2022MuKEA: Multimodal Knowledge Extraction and Accumulation for Knowledge-based Visual Question Answeringcited by 36
  34. 2021Probabilistic Modeling of Semantic Ambiguity for Scene Graph Generationcited by 35
  35. 2021Dual-Key Multimodal Backdoors for Visual Question Answeringcited by 35
  36. 2022RU-Net: Regularized Unrolling Network for Scene Graph Generationcited by 28
  37. 2022Fine-Grained Predicates Learning for Scene Graph Generationcited by 27
  38. 2019On Exploring Undetermined Relationships for Visual Relationship Detectioncited by 26

50 most cited of 57 All 57 (JSON)

Built on by

Cited by year

Citing papers per year: 2023: 4; 2024: 1.

5 papers.

  1. 2023HiLo: Exploiting High Low Frequency Relations for Unbiased Panoptic Scene Graph Generationcited by 19
  2. 2023Compositional Feature Augmentation for Unbiased Scene Graph Generationcited by 15
  3. 2023VLPrompt: Vision-Language Prompting for Panoptic Scene Graph Generationcited by 12
  4. 2024AUG: A New Dataset and An Efficient Model for Aerial Image Urban Scene Graph Generationcited by 5
  5. 2023CREPE: Learnable Prompting With CLIP Improves Visual Relationship Predictioncited by 4