Research map / Probabilistic and scientific modeling / Treatment effects
DROID: Driver-Centric Risk Object IDentification
Chengxi Li, Stanley H. Chan, Yi‐Ting Chen
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 GCN interaction modeling over I3D+LSTM | driver behavior model · HDDS Test1 | perplexity: 1.00 -> 0.83 | Improved |
| add intention multi-head (ablation) | GCN · HDDS Test1 | perplexity: 0.83 -> 0.72 | Improved |
| add TRN head with intention-initialized hidden state (ablation) | GCN · HDDS Test1 | perplexity: 0.83 -> 0.69 | Improved |
| add data augmentation via intervention (ablation) | GCN + TRN Head · HDDS Test1 | perplexity: 0.69 -> 0.37 | Improved |
| add Ego-Thing graph to Ego-Stuff graph (ablation) | GCN + TRN Head · HDDS Test1 | perplexity: 0.74 -> 0.69 | Improved |
| add Ego-Stuff graph to Ego-Thing graph (ablation) | GCN + TRN Head · HDDS Test1 | perplexity: 0.80 -> 0.69 | Improved |
| add 3D spatial relation to appearance relation (ablation) | GCN + TRN Head · HDDS Test1 | perplexity: 0.73 -> 0.69 | Improved |
| add intention head to GCN (ablation) | GCN · HDDS Test1 | driver intention mAP: 36.41 -> 36.31 | No clear effect |
| add data augmentation (ablation) | GCN + TRN Head · DROID Crossing Vehicle | mAcc: 29.0 -> 32.5 | Improved |
| add data augmentation (ablation) | GCN + TRN Head · DROID Congestion | mAcc: 52.2 -> 57.5 | Improved |
| add data augmentation (ablation) | GCN + TRN Head · DROID Crossing Pedestrian | mAcc: 13.2 -> 12.9 | Hurt |
| use causal intervention instead of correlation (attention) for risk object selection | GCN + TRN Head · DROID Parked Vehicle | mAcc: 18.7 -> 28.4 | Improved |
Closest papers by meaning
- Abductive Ego-View Accident Video Understanding for Safe Driving Perception (CVPR 2024 · similarity 0.84)
- Progressive Risk Estimation for Accident Anticipation (NeurIPS 2026 · similarity 0.84)
- On-Road Object Importance Estimation: A New Dataset and A Model with Multi-Fold Top-Down Guidance (NeurIPS 2024 · similarity 0.83)
- Dual-View Interaction-Aware Lane Change Prediction for Autonomous Driving (AAAI 2025 · similarity 0.83)
- DME-Driver: Integrating Human Decision Logic and 3D Scene Perception in Autonomous Driving (AAAI 2025 · similarity 0.82)
- DeepAccident: A Motion and Accident Prediction Benchmark for V2X Autonomous Driving (AAAI 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.
