atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Anticipating Traffic Accidents With Adaptive Loss and Large-Scale Incident DB

CVPR 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 136 citations (OpenAlex)

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 changedWhereMeasuredEffect
Replace Chan16's EL with AdaLEA (ablation)Chan16 · NIDB risk-factor anticipationmAP: +1.4Helped
Replace Chan16's EL with AdaLEA (ablation)Chan16 · NIDB risk-factor anticipationATTC: +0.28Helped
Replace LSTM with QRNN (ablation)Chan16 · DAD and NIDBmAP: 48.1 -> 51.7; 92.5 -> 94.2Helped
Replace LSTM with QRNN (ablation)Chan16 · DAD and NIDBATTC: 1.34 -> 3.02; 2.45 -> 2.85Helped
Use LEA instead of EL (ablation)QRNN risk anticipation model · DAD and NIDBATTC: 3.02 -> 3.22; 2.85 -> 4.67Helped
Use AdaLEA instead of LEA (ablation)QRNN risk anticipation model · DAD and NIDBATTC: 3.22 -> 3.43; 4.67 -> 4.72Helped
Add NIDB-pretrained global features (ablation)QRNN + AdaLEA · DAD and NIDBmAP: 52.3 -> 53.2; 96.3 -> 99.1Helped
Add NIDB-pretrained global features (ablation)QRNN + AdaLEA · DAD and NIDBATTC: 3.43 -> 3.44; 4.72 -> 4.81Helped
Use adaptive early-anticipation weighting rather than EL (ablation)Risk-factor anticipation model · NIDBATTCHelped

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy