Research map / Vision-language and multimodal / Cross-modal learning
MM-Fi: Multi-Modal Non-Intrusive 4D Human Dataset for Versatile Wireless Sensing
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| fuse RGB with LiDAR (I+L) instead of LiDAR alone | LiDAR HPE · MM-Fi P1 S1 | MPJPE (mm): 98.1 -> 94.3 | Helped |
| fuse LiDAR and WiFi with mmWave (R+L+W) instead of mmWave alone | mmWave HPE · MM-Fi P3 S1 | PA-MPJPE (mm): 57.3 -> 42.7 | Helped |
| add RGB to R+L+W fusion (R+L+W+I) (ablation) | R+L+W fusion · MM-Fi P1 S1 | MPJPE (mm): 74.1 -> 72.9 | Mixed |
| use cross-environment split instead of random split for LiDAR | LiDAR HPE · MM-Fi P3 | MPJPE (mm): 92.5 -> 303.8 | Hurt |
| use cross-environment split instead of random split for mmWave | mmWave HPE · MM-Fi P3 | MPJPE (mm): 117.0 -> 161.6 | Hurt |
| use cross-subject split instead of random split for WiFi | WiFi HPE · MM-Fi P1 | MPJPE (mm): 186.9 -> 222.3 | Hurt |
| retrain RGB model on MM-Fi instead of using Human3.6M pretrained (ablation) | VideoPose3D · MM-Fi P3 S1 | MPJPE (mm): 279.0 -> 60.5 | Helped |
| use RGB skeletons vs LiDAR skeletons for action recognition with AGCN | AGCN · MM-Fi action recognition | Top-1 accuracy (%): 54.44 -> 87.78 | Helped |
| use CTRGCN instead of AGCN on LiDAR skeletons | skeleton action recognition · MM-Fi action recognition | Top-1 accuracy (%): 54.44 -> 35.97 | Hurt |
Closest papers by meaning
- X-Fi: A Modality-Invariant Foundation Model for Multimodal Human Sensing (ICLR 2025 · similarity 0.88)
- MVDoppler-Pose: Multi-Modal Multi-View mmWave Sensing for Long-Distance Self-Occluded Human Walking Pose Estimation (CVPR 2025 · similarity 0.84)
- Ego4o: Egocentric Human Motion Capture and Understanding from Multi-Modal Input (CVPR 2025 · similarity 0.82)
- MITO: A Millimeter-Wave Dataset and Simulator for Non-Line-of-Sight Perception (NeurIPS 2026 · similarity 0.82)
- CSI-Bench: A Large-Scale In-the-Wild Dataset for Multi-task WiFi Sensing (NeurIPS 2025 · similarity 0.82)
- RELI11D: A Comprehensive Multimodal Human Motion Dataset and Method (CVPR 2024 · similarity 0.81)
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.