DensePose From WiFi
2022/12/31 by Jiaqi Geng, Dong Huang, Geng, Jiaqi +3 · 14 voices · 2 citations
Computer Science · Engineering · #Energy Efficient Wireless Sensor Networks #Indoor and Outdoor Localization Technologies #IoT-based Smart Home Systems #cs.CV
paper · pdf · doi:10.48550/arxiv.2301.00250
openalex publication_date 2022/12/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Abstract
Advances in computer vision and machine learning techniques have led to significant development in 2D and 3D human pose estimation from RGB cameras, LiDAR, and radars. However, human pose estimation from images is adversely affected by occlusion and lighting, which are common in many scenarios of interest. Radar and LiDAR technologies, on the other hand, need specialized hardware that is expensive and power-intensive. Furthermore, placing these sensors in non-public areas raises significant privacy concerns. To address these limitations, recent research has explored the use of WiFi antennas (1D sensors) for body segmentation and key-point body detection. This paper further expands on the use of the WiFi signal in combination with deep learning architectures, commonly used in computer vision, to estimate dense human pose correspondence. We developed a deep neural network that maps the phase and amplitude of WiFi signals to UV coordinates within 24 human regions. The results of the study reveal that our model can estimate the dense pose of multiple subjects, with comparable performance to image-based approaches, by utilizing WiFi signals as the only input. This paves the way for low-cost, broadly accessible, and privacy-preserving algorithms for human sensing.
Cited by
Discussions
- DensePose from WiFi [hn, 176 points, 44 comments]
- DensePose from WiFi Sensing: Full Body Tracking [hn, 4 points, 1 comments]
- I think, it's still a little funny you could theoretically track someone through wifi signals, even if it is pretty variable on metal surroundings arxiv.org/pdf/2301.00250 [bsky, 3 points, 0 comments]
- Рутерите са толкова умни, че вече може да картографират къде си по интерференчната картина, която сам създаваш. arxiv.org/pdf/2301.00250 [bsky, 2 points, 1 comments]
- cameras not even required! DensePost from WiFi (CMU, 2023) > We developed a deep neural network that maps the phase and amplitude of WiFi signals to UV coordinates within 24 human regions. arxiv.org/p [bsky, 2 points, 0 comments]
- Accurate Pose Estimation via WiFi Signals and ML [hn, 1 points, 0 comments]
- arxiv.org/pdf/2301.00250 I just found out about this today via a RSS feed for VR and Full Body Tracking. [bsky, 1 points, 0 comments]
- found this again, reminder that we do in fact have tiny radars inside our homes arxiv.org/abs/2301.00250 [bsky, 0 points, 0 comments]
- It’s been researched for a while now. arxiv.org/abs/2301.00250 [bsky, 0 points, 0 comments]
- what if, oh, I don't know, the government installed a little system in every single house (and gave everyone an always-on portable version) that could track your movements and map everything oh wait h [bsky, 0 points, 1 comments]
- DensePose from WiFi [bsky, 0 points, 0 comments]
- C'est basé sur un papier de recherche de Carnegie Mellon publié en 2023 mais de là à en faire un projet open-source fonctionnel... arxiv.org/abs/2301.00250 [bsky, 0 points, 1 comments]
- This from Dec 2022 (from an unrelated team at CMU) gives you some idea of the state of the art. arxiv.org/abs/2301.00250 [bsky, 0 points, 0 comments]
- isn't that a follow-up on Meta's DensePose? arxiv.org/abs/2301.00250 [bsky, 0 points, 1 comments]
Related