Research map / Video / Action recognition
Masked Video and Body-worn IMU Autoencoder for Egocentric Action Recognition
Mingfang Zhang, Yifei Huang, Ruicong Liu, Yoichi Sato
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 |
|---|---|---|---|
| initialize video encoder with VideoMAE pretrained on Kinetics-400 (ablation) | video encoder · CMU-MMAC action recognition (video only) | accuracy: 77.65 -> 85.07 | Improved |
| initialize video encoder with VideoMAE in IMU-Video model (ablation) | IMU-Video model · CMU-MMAC action recognition | accuracy: 79.89 -> 87.96 | Improved |
| initialize IMU encoder with ImageMAE trained on ImageNet (ablation) | IMU encoder · CMU-MMAC action recognition (IMU only) | accuracy: 31.65 -> 31.68 | No clear effect |
| use graph masking ratio 0.5 instead of 0.25 (ablation) | IMU pretraining · CMU-MMAC action recognition (IMU only) | accuracy: 29.42 -> 31.68 | Improved |
| use graph masking ratio 0.5 instead of 0.75 (ablation) | IMU pretraining · CMU-MMAC action recognition (IMU only) | accuracy: 31.45 -> 31.68 | Improved |
| use graph masking ratio 0.5 instead of 0.75 for IMU-Video (ablation) | IMU-Video pretraining · CMU-MMAC action recognition | accuracy: 86.26 -> 87.96 | Improved |
Closest papers by meaning
- Ego-Only: Egocentric Action Detection without Exocentric Transferring (ICCV 2023 · similarity 0.87)
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.84)
- Retrieval-Augmented Egocentric Video Captioning (CVPR 2024 · similarity 0.83)
- Whole-Body Conditioned Egocentric Video Prediction (NeurIPS 2025 · similarity 0.83)
- Estimating Ego-Body Pose from Doubly Sparse Egocentric Video Data (NeurIPS 2024 · similarity 0.83)
- SoundingActions: Learning How Actions Sound from Narrated Egocentric Videos (CVPR 2024 · similarity 0.82)
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.
