人工智能安全竞赛（AISC）
密码学与安全
2022-12-08 v1 人工智能
计算机视觉与模式识别
机器学习
摘要
人工智能（AI）的安全性是朝向安全、可靠和可信 AI 系统的重要研究领域。为加速 AI 安全研究，人工智能安全竞赛（AISC）由中关村实验室、国家工业信息安全发展研究中心、清华大学人工智能研究院和 RealAI 组织，作为中关村国际前沿技术创新大赛的一部分（https://www.zgc-aisc.com/en）。该竞赛包含三个赛道，包括 Deepfake 安全竞赛、自动驾驶安全竞赛和人脸识别安全竞赛。本报告将介绍这三个赛道的竞赛规则以及各赛道排名靠前团队的解决方案。
引用
@article{arxiv.2212.03412,
title = {Artificial Intelligence Security Competition (AISC)},
author = {Yinpeng Dong and Peng Chen and Senyou Deng and Lianji L and Yi Sun and Hanyu Zhao and Jiaxing Li and Yunteng Tan and Xinyu Liu and Yangyi Dong and Enhui Xu and Jincai Xu and Shu Xu and Xuelin Fu and Changfeng Sun and Haoliang Han and Xuchong Zhang and Shen Chen and Zhimin Sun and Junyi Cao and Taiping Yao and Shouhong Ding and Yu Wu and Jian Lin and Tianpeng Wu and Ye Wang and Yu Fu and Lin Feng and Kangkang Gao and Zeyu Liu and Yuanzhe Pang and Chengqi Duan and Huipeng Zhou and Yajie Wang and Yuhang Zhao and Shangbo Wu and Haoran Lyu and Zhiyu Lin and Yifei Gao and Shuang Li and Haonan Wang and Jitao Sang and Chen Ma and Junhao Zheng and Yijia Li and Chao Shen and Chenhao Lin and Zhichao Cui and Guoshuai Liu and Huafeng Shi and Kun Hu and Mengxin Zhang},
journal= {arXiv preprint arXiv:2212.03412},
year = {2022}
}
备注
Technical report of AISC