Research map / 3D vision / Human pose estimation
EventHands: Real-Time Neural 3D Hand Pose Estimation From an Event Stream
Viktor Rudnev, Vladislav Golyanik, Jiayi Wang, Hans-Peter Seidel, Franziska Mueller, Mohamed Elgharib, Christian Theobalt
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add data augmentation (ablation) | EventHands · EventHands real test | 2D-AUCp: 0.70 -> 0.77 | Improved |
| add data augmentation (ablation) | EventHands · EventHands synthetic test | 2D-AUCp: 0.88 -> 0.88 | No clear effect |
| add Kalman temporal filtering (ablation) | EventHands · EventHands real test | 2D-AUCp: 0.75 -> 0.77 | Improved |
| add Kalman temporal filtering (ablation) | EventHands · EventHands synthetic test | 2D-AUCp: 0.89 -> 0.88 | No clear effect |
| use LNES instead of EOI at 100ms window (ablation) | event representation · EventHands real test | 2D-AUCp: 0.56 -> 0.77 | Improved |
| use LNES instead of ECI at 100ms window (ablation) | event representation · EventHands real test | 2D-AUCp: 0.52 -> 0.77 | Improved |
| use LNES instead of ECI-S at 100ms window (ablation) | event representation · EventHands real test | 2D-AUCp: 0.56 -> 0.77 | Improved |
| use LNES instead of EOI at 33ms window (ablation) | event representation · EventHands real test | 2D-AUCp: 0.70 -> 0.72 | Improved |
| lengthen LNES window from 100ms to 300ms (ablation) | LNES · EventHands real test | 2D-AUCp: 0.77 -> 0.72 | Hurt |
| lengthen LNES window from 33ms to 100ms (ablation) | LNES · EventHands real test | 2D-AUCp: 0.72 -> 0.77 | Improved |
| lengthen EOI window from 33ms to 100ms (ablation) | EOI · EventHands real test | 2D-AUCp: 0.70 -> 0.56 | Hurt |
| lengthen ECI window from 33ms to 100ms (ablation) | ECI · EventHands real test | 2D-AUCp: 0.69 -> 0.52 | Hurt |
Later papers that compared with this method
3 later accepted papers report EventHands as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Complementing Event Streams and RGB Frames for Hand Mesh Reconstruction (CVPR 2024 · better on 5 of 5 shared results)
- EvHandPose: Event-Based 3D Hand Pose Estimation With Sparse Supervision (TPAMI 2024 · better on 5 of 5 shared results)
- 3D Hand Sequence Recovery from Real Blurry Images and Event Stream (ECCV 2024 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- EvHandPose: Event-Based 3D Hand Pose Estimation With Sparse Supervision (TPAMI 2024 · similarity 0.92)
- SeqHAND: RGB-Sequence-Based 3D Hand Pose and Shape Estimation (ECCV 2020 · similarity 0.88)
- Monocular Real-Time Hand Shape and Motion Capture Using Multi-Modal Data (CVPR 2020 · similarity 0.88)
- EventCap: Monocular 3D Capture of High-Speed Human Motions Using an Event Camera (CVPR 2020 · similarity 0.87)
- Complementing Event Streams and RGB Frames for Hand Mesh Reconstruction (CVPR 2024 · similarity 0.87)
- EventEgo3D: 3D Human Motion Capture from Egocentric Event Streams (CVPR 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
