mlsanityData to 14 Oct 2026

Focus on Low-Resolution Information: Multi-Granular Information-Lossless Model for Low-Resolution Human Pose Estimation

arXiv 2405.12247 ↗First posted 19 May 2024cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

In real-world applications of human pose estimation, low-resolution input images are frequently encountered when the performance of the image acquisition equipment is limited or the shooting distance is too far. However, existing state-of-the-art models for human pose estimation perform poorly on low-resolution images. One key reason is the presence of downsampling layers in these models, e.g., strided convolutions and pooling layers. It further reduces the already insufficient image information. Another key reason is that the body skeleton and human kinematic information are not fully utilized. In this work, we propose a Multi-Granular Information-Lossless (MGIL) model to replace the downsampling layers to address the above issues. Specifically, MGIL employs a Fine-grained Lossless Information Extraction (FLIE) module, which can prevent the loss of local information. Furthermore, we design a Coarse-grained Information Interaction (CII) module to adequately leverage human body structural information. To efficiently fuse cross-granular information and thoroughly exploit the relationships among keypoints, we further introduce a Multi-Granular Adaptive Fusion (MGAF) mechanism. The mechanism assigns weights to features of different granularities based on the content of the image. The model is effective, flexible, and universal. We show its potential in various vision tasks with comprehensive experiments. It outperforms the SOTA methods by 7.7 mAP on COCO and performs well with different input resolutions, different backbones, and different vision tasks. The code is provided in supplementary material.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 51 of its 75 references that are in the map. First posted on arXiv 19 May 2024.

Builds on

51 of 75 references.

References · 51

  1. 2017Attention Is All You Needcited by 60,361
  2. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  3. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  4. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,852
  5. 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
  6. 2019Objects as Pointscited by 1,175
  7. 2016Realtime Multi-Person 2D Pose Estimation using Part Affinity Fieldscited by 1,064
  8. 2016Deep Networks with Stochastic Depthcited by 979
  9. 2016Stacked Hourglass Networks for Human Pose Estimationcited by 915
  10. 2018Simple Baselines for Human Pose Estimation and Trackingcited by 502
  11. 2021Nystr"omformer: A Nystr"om-Based Algorithm for Approximating Self-Attentioncited by 471
  12. 2017Cascaded Pyramid Network for Multi-Person Pose Estimationcited by 341
38 more
  1. 2016RMPE: Regional Multi-person Pose Estimationcited by 340
  2. 2016Convolutional Pose Machinescited by 325
  3. 2023Effective Whole-body Pose Estimation with Two-stages Distillationcited by 311
  4. 2013DeepPose: Human Pose Estimation via Deep Neural Networkscited by 305
  5. 2019Semantic Graph Convolutional Networks for 3D Human Pose Regressioncited by 204
  6. 2019Distribution-Aware Coordinate Representation for Human Pose Estimationcited by 165
  7. 2017Towards Accurate Multi-person Pose Estimation in the Wildcited by 131
  8. 2021Human Pose Regression with Residual Log-likelihood Estimationcited by 131
  9. 2014Joint Training of a Convolutional Network and a Graphical Model for Human Pose Estimationcited by 129
  10. 2015Human Pose Estimation with Iterative Error Feedbackcited by 121
  11. 2021Bottom-Up Human Pose Estimation Via Disentangled Keypoint Regressioncited by 112
  12. 2016Resnet in Resnet: Generalizing Residual Architecturescited by 110
  13. 2018Pose2Seg: Detection Free Human Instance Segmentationcited by 107
  14. 2021TokenPose: Learning Keypoint Tokens for Human Pose Estimationcited by 107
  15. 2021Pose Recognition with Cascade Transformerscited by 99
  16. 2021Graph Stacked Hourglass Networks for 3D Human Pose Estimationcited by 95
  17. 2020SRNet: Improving Generalization in 3D Human Pose Estimation with a Split-and-Recombine Approachcited by 94
  18. 2019The Devil is in the Details: Delving into Unbiased Data Processing for Human Pose Estimationcited by 79
  19. 2018Fast Human Pose Estimationcited by 68
  20. 2023Human-Art: A Versatile Human-Centric Dataset Bridging Natural and Artificial Scenescited by 68
  21. 2019Rethinking on Multi-Stage Networks for Human Pose Estimationcited by 58
  22. 2017Learning Feature Pyramids for Human Pose Estimationcited by 57
  23. 2018Low-resolution Face Recognition in the Wild via Selective Knowledge Distillationcited by 55
  24. 2016Human pose estimation via Convolutional Part Heatmap Regressioncited by 51
  25. 2019DirectPose: Direct End-to-End Multi-Person Pose Estimationcited by 49
  26. 2020UniPose: Unified Human Pose Estimation in Single Images and Videoscited by 49
  27. 2023Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimationcited by 48
  28. 2019Multi-Person Pose Estimation with Enhanced Channel-wise and Spatial Informationcited by 35
  29. 2019Mixture Dense Regression for Object Detection and Human Pose Estimationcited by 29
  30. 2022No More Strided Convolutions or Pooling: A New CNN Building Block for Low-Resolution Images and Small Objectscited by 27
  31. 2021Multi-Scale Aligned Distillation for Low-Resolution Detectioncited by 24
  32. 2015Robust Optimization for Deep Regressioncited by 22
  33. 2022Teaching Where to Look: Attention Similarity Knowledge Distillation for Low Resolution Face Recognitioncited by 22
  34. 2023DistilPose: Tokenized Pose Regression with Heatmap Distillationcited by 12
  35. 2016Fine-to-coarse Knowledge Transfer For Low-Res Image Classificationcited by 11
  36. 2022WaveMix: Resource-efficient Token Mixing for Imagescited by 9
  37. 2018Stochastic Downsampling for Cost-Adjustable Inference and Improved Regularization in Convolutional Networkscited by 7
  38. 2020Learning to Learn Parameterized Classification Networks for Scalable Input Imagescited by 6

50 most cited of 51 All 51 (JSON)

Built on by

Not cited yet.