Research map / 3D vision / 3D generation
EgoTraj: Real-World Egocentric Human Trajectory
Ahmad Yehia, Abduallah Mohamed, Tianyi Wang, Kun Qian, Jiseop Byeon, Junfeng Jiao, Christian Claudel
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 gaze (u,v) to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.19 -> 0.15 | Improved |
| add scene segmentation to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.19 -> 0.16 | Improved |
| add body pose of nearby people to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.19 -> 0.17 | Improved |
| add center points of nearby people (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | L1 head rotation: 0.69 -> 0.79 | Hurt |
| add relative depth to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.19 -> 0.18 | Improved |
| add gaze on top of pose (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.17 -> 0.12 | Improved |
| add pose, segmentation and gaze to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | ADE: 0.19 -> 0.12 | Improved |
| add pose, segmentation and gaze to ego-motion (ablation) | CXA-Transformer · EgoTraj test split (modality ablation) | L1 head rotation: 0.69 -> 0.58 | Improved |
| add pose, segmentation and gaze (waypoint held-out split) (ablation) | CXA-Transformer · EgoTraj waypoint held-out (n=10) | ADE: 0.21 -> 0.14 | Improved |
| add pose, segmentation and gaze (unfamiliar participants split) (ablation) | CXA-Transformer · EgoTraj unfamiliar participants (n=8) | ADE: 0.23 -> 0.14 | Improved |
| use EgoCast instead of CXA-Transformer | EgoCast · EgoTraj test split | ADE: 0.19 -> 0.16 | Mixed |
| use learned Transformer instead of constant velocity | M_Transformer · EgoTraj test split | ADE: 0.24 -> 0.20 | Improved |
Closest papers by meaning
- EgoCogNav: Cognition-aware Human Egocentric Navigation (ECCV 2026 · similarity 0.90)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · similarity 0.89)
- Gaze Beyond the Frame: Forecasting Egocentric 3D Visual Span (NeurIPS 2025 · similarity 0.88)
- EgoSim: An Egocentric Multi-view Simulator and Real Dataset for Body-worn Cameras during Motion and Activity (NeurIPS 2024 · similarity 0.88)
- Leveraging Driver Field-of-View for Multimodal Ego-Trajectory Prediction (ICLR 2025 · similarity 0.87)
- RoboSense: Large-scale Dataset and Benchmark for Egocentric Robot Perception and Navigation in Crowded and Unstructured Environments (CVPR 2025 · similarity 0.86)
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.
