Risky Object Localization (ROL) in a Driving Scene Dataset

By Muhammad Monjurul Karim, Drew Racz, Gary Liu, Ziming Li, Yanguang Gong, Michael Incardona, William Li, Ruwen Qin, Zhaozheng Yin

Download

• Dataset : Download
• CNN Feature : Download

Description

For detailed information about the dataset, kindly refer to our paper. To train and test using this dataset, please download the mentioned extracted CNN features from the Download section.

Code for risky object localization in a driving scene

Please refer to AM-Net Official repository for the details of the Code for risky object localization with AM-Net.

Prediction Demo

Media1.mp4

Citation

Please cite our paper if you find the dataset useful.

@ARTICLE{karim_am_net2023,
  author={Karim, Muhammad Monjurul and Yin, Zhaozheng and Qin, Ruwen},
  journal={IEEE Transactions on Intelligent Vehicles}, 
  title={An Attention-guided Multistream Feature Fusion Network for Early Localization of Risky Traffic Agents in Driving Videos}, 
  year={2023},
  volume={},
  number={},
  pages={1-12},
  doi={10.1109/TIV.2023.3275543}}