Research map / Large language models / AI and society
trajdata: A Unified Interface to Multiple Human Trajectory Datasets
Boris Ivanovic, Guanyu Song, Igor Gilitschenski, Marco Pavone
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 |
|---|---|---|---|
| use drone-based collection instead of on-vehicle sensors | annotation self-consistency · INTERACTION vs AV datasets | collision, off-road, harsh acceleration rates | Improved |
| use autolabeling (Waymo) instead of human annotation (nuScenes) | annotation pipeline · vehicle collision rate | collision rate | No clear effect |
| use onboard perception annotations (Lyft) instead of human/autolabels | annotation pipeline · agent detection duration | observation duration | Hurt |
| use axis-aligned boxes (SDD) instead of oriented boxes | annotation · SDD self-consistency | collision rate | Hurt |
| compare pedestrian datasets against AV datasets for agent density | dataset choice · agent density per scene | agent density | Improved |
| compare Waymo local map crops against full-neighborhood maps | map format · storage | storage requirement | Hurt |
| use nuScenes maps away from main roads | map coverage · nuScenes annotation alignment | map alignment | Hurt |
Closest papers by meaning
- Pedestrian Trajectory Prediction with Missing Data: Datasets, Imputation, and Benchmarking (NeurIPS 2024 · similarity 0.92)
- Towards Predicting Any Human Trajectory In Context (NeurIPS 2025 · similarity 0.88)
- Social-Transmotion: Promptable Human Trajectory Prediction (ICLR 2024 · similarity 0.87)
- TrajMamba: An Efficient and Semantic-rich Vehicle Trajectory Pre-training Model (NeurIPS 2025 · similarity 0.87)
- GigaTraj: Predicting Long-term Trajectories of Hundreds of Pedestrians in Gigapixel Complex Scenes (CVPR 2024 · similarity 0.87)
- GTR: A General, Multi-View, and Dynamic Framework for Trajectory Representation Learning (ICML 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.
