mlsanityData to 7 Oct 2026

Learning Cross-Image Object Semantic Relation in Transformer for Few-Shot Fine-Grained Image Classification

arXiv 2207.00784 ↗First posted 2 Jul 2022cs.CVcited by 9

Not on the map. Why?

Abstract (arXiv)

Few-shot fine-grained learning aims to classify a query image into one of a set of support categories with fine-grained differences. Although learning different objects' local differences via Deep Neural Networks has achieved success, how to exploit the query-support cross-image object semantic relations in Transformer-based architecture remains under-explored in the few-shot fine-grained scenario. In this work, we propose a Transformer-based double-helix model, namely HelixFormer, to achieve the cross-image object semantic relation mining in a bidirectional and symmetrical manner. The HelixFormer consists of two steps: 1) Relation Mining Process (RMP) across different branches, and 2) Representation Enhancement Process (REP) within each individual branch. By the designed RMP, each branch can extract fine-grained object-level Cross-image Semantic Relation Maps (CSRMs) using information from the other branch, ensuring better cross-image interaction in semantically related local object regions. Further, with the aid of CSRMs, the developed REP can strengthen the extracted features for those discovered semantically-related local regions in each branch, boosting the model's ability to distinguish subtle feature differences of fine-grained objects. Extensive experiments conducted on five public fine-grained benchmarks demonstrate that HelixFormer can effectively enhance the cross-image object semantic relation matching for recognizing fine-grained objects, achieving much better performance over most state-of-the-art methods under 1-shot and 5-shot scenarios. Our code is available at: https://github.com/JiakangYuan/HelixFormer

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 52 of its 74 references that are in the map. First posted on arXiv 2 Jul 2022.

Builds on

52 of 74 references.

References · 52

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  4. 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
  5. 2017Squeeze-and-Excitation Networkscited by 4,763
  6. 2015SSD: Single Shot MultiBox Detectorcited by 3,125
  7. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  8. 2016Matching Networks for One Shot Learningcited by 2,246
  9. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  10. 2013Fine-Grained Visual Classification of Aircraftcited by 1,701
  11. 2017Learning to Compare: Relation Network for Few-Shot Learningcited by 1,139
  12. 2021CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classificationcited by 806
38 more
  1. 2018CCNet: Criss-Cross Attention for Semantic Segmentationcited by 717
  2. 2019A Closer Look at Few-shot Classificationcited by 679
  3. 2018Learning to Reweight Examples for Robust Deep Learningcited by 670
  4. 2021Twins: Revisiting the Design of Spatial Attention in Vision Transformerscited by 565
  5. 2018Meta-Learning with Latent Embedding Optimizationcited by 476
  6. 2018TADAM: Task dependent adaptive metric for improved few-shot learningcited by 408
  7. 2017Few-Shot Learning with Graph Neural Networkscited by 324
  8. 2018Meta-Transfer Learning for Few-Shot Learningcited by 306
  9. 2018Few-Shot Learning via Embedding Adaptation with Set-to-Set Functionscited by 293
  10. 2019Cross Attention Network for Few-shot Classificationcited by 218
  11. 2018Low-Shot Learning from Imaginary Datacited by 181
  12. 2020CrossTransformers: spatially-aware few-shot transfercited by 174
  13. 2021Free Lunch for Few-shot Learning: Distribution Calibrationcited by 165
  14. 2021HRFormer: High-Resolution Transformer for Dense Predictioncited by 149
  15. 2021CrossFormer: A Versatile Vision Transformer Hinging on Cross-scale Attentioncited by 139
  16. 2020Negative Margin Matters: Understanding Margin in Few-shot Classificationcited by 129
  17. 2019Revisiting Local Descriptor based Image-to-Class Measure for Few-shot Learningcited by 121
  18. 2020Few-Shot Classification with Feature Map Reconstruction Networkscited by 116
  19. 2020Adversarial Feature Hallucination Networks for Few-Shot Learningcited by 110
  20. 2019Meta-Learning without Memorizationcited by 92
  21. 2021CAT: Cross Attention in Vision Transformercited by 92
  22. 2020Learning Attentive Pairwise Interaction for Fine-Grained Classificationcited by 84
  23. 2020Meta-Learning with Adaptive Hyperparameterscited by 79
  24. 2019Improving Generalization in Meta Reinforcement Learning using Learned Objectivescited by 68
  25. 2020Instance Credibility Inference for Few-Shot Learningcited by 57
  26. 2019Image Deformation Meta-Networks for One-Shot Learningcited by 56
  27. 2019Learning to Balance: Bayesian Meta-Learning for Imbalanced and Out-of-distribution Taskscited by 48
  28. 2021Prototypical Cross-Attention Networks for Multiple Object Tracking and Segmentationcited by 43
  29. 2019Low-Rank Pairwise Alignment Bilinear Network For Few-Shot Fine-Grained Image Classificationcited by 36
  30. 2020Revisiting Pose-Normalization for Fine-Grained Few-Shot Recognitioncited by 30
  31. 2020BSNet: Bi-Similarity Network for Few-shot Fine-grained Image Classificationcited by 30
  32. 2020Restoring Negative Information in Few-Shot Object Detectioncited by 28
  33. 2019PARN: Position-Aware Relation Networks for Few-Shot Learningcited by 27
  34. 2017Few-shot Autoregressive Density Estimation: Towards Learning to Learn Distributionscited by 24
  35. 2021ZiGAN: Fine-grained Chinese Calligraphy Font Generation via a Few-shot Style Transfer Approachcited by 20
  36. 2020TOAN: Target-Oriented Alignment Network for Fine-Grained Image Categorization with Few Labeled Samplescited by 13
  37. 2021Object-aware Long-short-range Spatial Alignment for Few-Shot Fine-Grained Image Classificationcited by 8
  38. 2020MetaPerturb: Transferable Regularizer for Heterogeneous Tasks and Architecturescited by 7

50 most cited of 52 All 52 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 3; 2024: 0; 2025: 1; 2026: 3.

9 papers.

  1. 2023Robust Saliency-Aware Distillation for Few-shot Fine-grained Visual Recognitioncited by 6
  2. 2023Masked Cross-image Encoding for Few-shot Segmentationcited by 3
  3. 2023Multi-Content Interaction Network for Few-Shot Segmentationcited by 2
  4. 2026Adaptive receptive field-based spatial-frequency feature reconstruction network for fine-grained few-shot image classificationcited by 2
  5. 2022UniDA3D: Unified Domain Adaptive 3D Semantic Segmentation Pipelinecited by 1
  6. 2026Frequency-Enhanced Dual-Subspace Networks for Few-Shot Fine-Grained Image Classificationcited by 1
  7. 2022Fine-grained Few-shot Recognition by Deep Object Parsingcited by 0
  8. 2025Hierarchical Mask-Enhanced Dual Reconstruction Network for Few-Shot Fine-Grained Image Classificationcited by 0
  9. 2026Phase-Aware Spatial-Frequency Fusion for Few-Shot Fine-Grained Image Classificationcited by 0