mlsanityData to 14 Oct 2026

Deformable ConvNets v2: More Deformable, Better Results

arXiv 1811.11168 ↗First posted 27 Nov 2018cs.CVcited by 670

Not on the map. Why?

Abstract (arXiv)

The superior performance of Deformable Convolutional Networks arises from its ability to adapt to the geometric variations of objects. Through an examination of its adaptive behavior, we observe that while the spatial support for its neural features conforms more closely than regular ConvNets to object structure, this support may nevertheless extend well beyond the region of interest, causing features to be influenced by irrelevant image content. To address this problem, we present a reformulation of Deformable ConvNets that improves its ability to focus on pertinent image regions, through increased modeling power and stronger training. The modeling power is enhanced through a more comprehensive integration of deformable convolution within the network, and by introducing a modulation mechanism that expands the scope of deformation modeling. To effectively harness this enriched modeling capability, we guide network training via a proposed feature mimicking scheme that helps the network to learn features that reflect the object focus and classification power of R-CNN features. With the proposed contributions, this new version of Deformable ConvNets yields significant performance gains over the original model and produces leading results on the COCO benchmark for object detection and instance segmentation.

Key facts. Cited by 670 papers in the map (data through 14 Oct 2026). First posted on arXiv 27 Nov 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 172; 2022: 136; 2023: 125; 2024: 98; 2025: 85; 2026: 54.

670 papers.

Sort:

By area:

  1. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  2. 2024VMamba: Visual State Space Modelcited by 969
  3. 2021Open-vocabulary Object Detection via Vision and Language Knowledge Distillationcited by 906
  4. 2022BEVDepth: Acquisition of Reliable Depth for Multi-view 3D Object Detectioncited by 572
  5. 2022InternImage: Exploring Large-Scale Vision Foundation Models with Deformable Convolutionscited by 482
  6. 2021Dynamic Neural Networks: A Surveycited by 408
  7. 2021BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignmentcited by 342
  8. 2022Vision Transformer with Deformable Attentioncited by 295
  9. 2021Attention Mechanisms in Computer Vision: A Surveycited by 232
  10. 2021Contrastive Learning for Compact Single Image Dehazingcited by 219
  11. 2021TOOD: Task-aligned One-stage Object Detectioncited by 218
  12. 2021K-Net: Towards Unified Image Segmentationcited by 212
38 more
  1. 2022Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlationcited by 197
  2. 2023ProPainter: Improving Propagation and Transformer for Video Inpaintingcited by 197
  3. 2021Boundary IoU: Improving Object-Centric Image Segmentation Evaluationcited by 188
  4. 2021Dynamic Head: Unifying Object Detection Heads with Attentionscited by 187
  5. 2021OTA: Optimal Transport Assignment for Object Detectioncited by 178
  6. 2022Recurrent Video Restoration Transformer with Guided Deformable Attentioncited by 170
  7. 2021FSCE: Few-Shot Object Detection via Contrastive Proposal Encodingcited by 168
  8. 2021Instances as Queriescited by 158
  9. 2022Focal Sparse Convolutional Networks for 3D Object Detectioncited by 142
  10. 2021Delving into Localization Errors for Monocular 3D Object Detectioncited by 140
  11. 2021Probabilistic two-stage detectioncited by 130
  12. 2021OW-DETR: Open-world Detection Transformercited by 129
  13. 2021Efficient DETR: Improving End-to-End Object Detector with Dense Priorcited by 128
  14. 2023Transformer-Based Visual Segmentation: A Surveycited by 124
  15. 2023ALIKED: A Lighter Keypoint and Descriptor Extraction Network via Deformable Transformationcited by 121
  16. 2021Deformable ProtoPNet: An Interpretable Image Classifier Using Deformable Prototypescited by 120
  17. 2021TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformerscited by 117
  18. 2022VQFR: Blind Face Restoration with Vector-Quantized Dictionary and Parallel Decodercited by 116
  19. 2022Towards An End-to-End Framework for Flow-Guided Video Inpaintingcited by 113
  20. 2022VideoINR: Learning Video Implicit Neural Representation for Continuous Space-Time Super-Resolutioncited by 101
  21. 2021Involution: Inverting the Inherence of Convolution for Visual Recognitioncited by 99
  22. 2022Group DETR: Fast DETR Training with Group-Wise One-to-Many Assignmentcited by 98
  23. 2023Large Selective Kernel Network for Remote Sensing Object Detectioncited by 98
  24. 2024Vision-RWKV: Efficient and Scalable Visual Perception with RWKV-Like Architecturescited by 98
  25. 2022Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusioncited by 94
  26. 2021Lite-HRNet: A Lightweight High-Resolution Networkcited by 91
  27. 2023TRACE: 5D Temporal Regression of Avatars with Dynamic Cameras in 3D Environmentscited by 91
  28. 2023DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignmentcited by 90
  29. 2021Aggregated Contextual Transformations for High-Resolution Image Inpaintingcited by 89
  30. 2022Rethinking Efficient Lane Detection via Curve Modelingcited by 88
  31. 2021FaPN: Feature-aligned Pyramid Network for Dense Image Predictioncited by 81
  32. 2021Fourier Contour Embedding for Arbitrary-Shaped Text Detectioncited by 80
  33. 2021MASA-SR: Matching Acceleration and Spatial Adaptation for Reference-Based Image Super-Resolutioncited by 80
  34. 2022AdaMixer: A Fast-Converging Query-Based Object Detectorcited by 79
  35. 2022Changer: Feature Interaction is What You Need for Change Detectioncited by 79
  36. 2023BEVHeight: A Robust Framework for Vision-based Roadside 3D Object Detectioncited by 75
  37. 2021Open-World Entity Segmentationcited by 74
  38. 2022LargeKernel3D: Scaling up Kernels in 3D Sparse CNNscited by 72

50 most cited of 670 All 670 (JSON)