Sparse and Privacy-enhanced Representation for Human Pose Estimation
Not on the map. Why?
Abstract (arXiv)
We propose a sparse and privacy-enhanced representation for Human Pose Estimation (HPE). Given a perspective camera, we use a proprietary motion vector sensor(MVS) to extract an edge image and a two-directional motion vector image at each time frame. Both edge and motion vector images are sparse and contain much less information (i.e., enhancing human privacy). We advocate that edge information is essential for HPE, and motion vectors complement edge information during fast movements. We propose a fusion network leveraging recent advances in sparse convolution used typically for 3D voxels to efficiently process our proposed sparse representation, which achieves about 13x speed-up and 96% reduction in FLOPs. We collect an in-house edge and motion vector dataset with 16 types of actions by 40 users using the proprietary MVS. Our method outperforms individual modalities using only edge or motion vector images. Finally, we validate the privacy-enhanced quality of our sparse representation through face recognition on CelebA (a large face dataset) and a user study on our in-house dataset.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 38 of its 55 references that are in the map. First posted on arXiv 18 Sep 2023.
Builds on
38 of 55 references.
References · 38
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
- 2014Microsoft COCO: Common Objects in Contextcited by 17,498
- 2017Mask R-CNNcited by 7,259
- 2014Deep Learning Face Attributes in the Wildcited by 3,679
- 2019Deep High-Resolution Representation Learning for Human Pose Estimationcited by 1,339
- 2016Realtime Multi-Person 2D Pose Estimation using Part Affinity Fieldscited by 1,061
- 2016Stacked Hourglass Networks for Human Pose Estimationcited by 913
- 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
- 2018Simple Baselines for Human Pose Estimation and Trackingcited by 502
- 2017Cascaded Pyramid Network for Multi-Person Pose Estimationcited by 341
- 2016OctNet: Learning Deep 3D Representations at High Resolutionscited by 340
26 more
- 2016RMPE: Regional Multi-person Pose Estimationcited by 340
- 2016Convolutional Pose Machinescited by 325
- 2013DeepPose: Human Pose Estimation via Deep Neural Networkscited by 305
- 2017Submanifold Sparse Convolutional Networkscited by 223
- 2018CrowdPose: Efficient Crowded Scenes Pose Estimation and A New Benchmarkcited by 214
- 2018SPLATNet: Sparse Lattice Networks for Point Cloud Processingcited by 158
- 2017Towards Accurate Multi-person Pose Estimation in the Wildcited by 131
- 2015DeepCut: Joint Subset Partition and Labeling for Multi Person Pose Estimationcited by 117
- 2016DeeperCut: A Deeper, Stronger, and Faster Multi-Person Pose Estimation Modelcited by 96
- 2016Vote3Deep: Fast Object Detection in 3D Point Clouds Using Efficient Convolutional Neural Networkscited by 72
- 2015Sparse 3D convolutional neural networkscited by 71
- 2018Learning to Anonymize Faces for Privacy Preserving Action Detectioncited by 63
- 2017Learning Feature Pyramids for Human Pose Estimationcited by 57
- 2018ArcFace: Additive Angular Margin Loss for Deep Face Recognitioncited by 56
- 2018Towards Privacy-Preserving Visual Recognition via Adversarial Training: A Pilot Studycited by 48
- 2023Explicit Box Detection Unifies End-to-End Multi-Person Pose Estimationcited by 48
- 2016Privacy-Preserving Human Activity Recognition from Extreme Low Resolutioncited by 43
- 2021EventHPE: Event-based 3D Human Pose and Shape Estimationcited by 42
- 2014Spatially-sparse convolutional neural networkscited by 33
- 2022DeltaCNN: End-to-End CNN Inference of Sparse Frame Differences in Videoscited by 18
- 2022PrivHAR: Recognizing Human Actions From Privacy-preserving Lenscited by 16
- 2015Combining Local Appearance and Holistic View: Dual-Source Deep Neural Networks for Human Pose Estimationcited by 14
- 2021In-Bed Human Pose Estimation from Unseen and Privacy-Preserving Image Domainscited by 7
- 20183D human pose estimation from depth maps using a deep combination of posescited by 6
- 2021Sketch-based Creativity Support Tools using Deep Learningcited by 3
- 2022Location-free Human Pose Estimationcited by 1
Built on by
Citing papers per year: 2023: 1; 2024: 0; 2025: 1.
2 papers.