mlsanityData to 14 Oct 2026

Distilling Token-Pruned Pose Transformer for 2D Human Pose Estimation

arXiv 2304.05548 ↗First posted 12 Apr 2023cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Human pose estimation has seen widespread use of transformer models in recent years. Pose transformers benefit from the self-attention map, which captures the correlation between human joint tokens and the image. However, training such models is computationally expensive. The recent token-Pruned Pose Transformer (PPT) solves this problem by pruning the background tokens of the image, which are usually less informative. However, although it improves efficiency, PPT inevitably leads to worse performance than TokenPose due to the pruning of tokens. To overcome this problem, we present a novel method called Distilling Pruned-Token Transformer for human pose estimation (DPPT). Our method leverages the output of a pre-trained TokenPose to supervise the learning process of PPT. We also establish connections between the internal structure of pose transformers and PPT, such as attention maps and joint features. Our experimental results on the MPII datasets show that our DPPT can significantly improve PCK compared to previous PPT models while still reducing computational complexity.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 50 of its 58 references that are in the map. First posted on arXiv 12 Apr 2023.

Builds on

50 of 58 references.

References · 50

  1. 2017Attention Is All You Needcited by 60,361
  2. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  4. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  5. 2020End-to-End Object Detection with Transformerscited by 7,414
  6. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  7. 2020Training data-efficient image transformers & distillation through attentioncited by 4,586
  8. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
  9. 2020Rethinking Semantic Segmentation from a Sequence-to-Sequence Perspective with Transformerscited by 1,369
  10. 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
  11. 2018Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognitioncited by 1,209
  12. 2021Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNetcited by 1,101
38 more
  1. 2016Stacked Hourglass Networks for Human Pose Estimationcited by 915
  2. 2021DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsificationcited by 864
  3. 2021Segmenter: Transformer for Semantic Segmentationcited by 855
  4. 2019Improved Knowledge Distillation via Teacher Assistantcited by 573
  5. 2018Simple Baselines for Human Pose Estimation and Trackingcited by 502
  6. 2019Be Your Own Teacher: Improve the Performance of Convolutional Neural Networks via Self Distillationcited by 497
  7. 2022ViTPose: Simple Vision Transformer Baselines for Human Pose Estimationcited by 456
  8. 2020End-to-End Human Pose and Mesh Reconstruction with Transformerscited by 418
  9. 2019Q-BERT: Hessian Based Ultra Low Precision Quantization of BERTcited by 359
  10. 2017Cascaded Pyramid Network for Multi-Person Pose Estimationcited by 341
  11. 2016Convolutional Pose Machinescited by 325
  12. 2019Variational Information Distillation for Knowledge Transfercited by 324
  13. 2018Learning Deep Representations with Probabilistic Knowledge Transfercited by 213
  14. 2021Chasing Sparsity in Vision Transformers: An End-to-End Explorationcited by 183
  15. 2021You Only Look at One Sequence: Rethinking Transformer in Vision through Object Detectioncited by 170
  16. 2014Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimationcited by 129
  17. 2020TransPose: Keypoint Localization via Transformercited by 126
  18. 2021TokenPose: Learning Keypoint Tokens for Human Pose Estimationcited by 107
  19. 2021Pose Recognition with Cascade Transformerscited by 99
  20. 2022Unified Visual Transformer Compressioncited by 79
  21. 2022Knowledge Distillation via the Target-aware Transformercited by 70
  22. 2018Fast Human Pose Estimationcited by 68
  23. 2017Multi-Context Attention for Human Pose Estimationcited by 66
  24. 2022DearKD: Data-Efficient Early Knowledge Distillation for Vision Transformerscited by 65
  25. 2021Online Knowledge Distillation for Efficient Pose Estimationcited by 55
  26. 2021AFTer-UNet: Axial Fusion Transformer UNet for Medical Image Segmentationcited by 48
  27. 2016Deep Learning on Lie Groups for Skeleton-based Action Recognitioncited by 42
  28. 2021TFPose: Direct Human Pose Estimation with Transformerscited by 42
  29. 2020Predicting Camera Viewpoint Improves Cross-dataset Generalization for 3D Human Pose Estimationcited by 36
  30. 2021TransFusion: Cross-view Fusion with Transformer for 3D Human Pose Estimationcited by 32
  31. 2022VAQF: Fully Automatic Software-Hardware Co-Design Framework for Low-Bit Vision Transformercited by 27
  32. 2022PPT: token-Pruned Pose Transformer for monocular and multi-view human pose estimationcited by 27
  33. 2022Peeling the Onion: Hierarchical Reduction of Data Redundancy for Efficient Vision Transformer Trainingcited by 24
  34. 2020Nonparametric Structure Regularization Machine for 2D Hand Pose Estimationcited by 12
  35. 2019Adaptive Graphical Model Network for 2D Handpose Estimationcited by 9
  36. 2020SIA-GCN: A Spatial Information Aware Graph Neural Network with 2D Convolutions for Hand Pose Estimationcited by 9
  37. 2022The Best of Both Worlds: Combining Model-based and Nonparametric Approaches for 3D Human Body Estimationcited by 8
  38. 2020Rotation-invariant Mixed Graphical Model Network for 2D Hand Pose Estimationcited by 6

Built on by

Not cited yet.