【论文阅读】关于自动驾驶的开环训练和评估，到底让模型学到了什么？

1. Ego status & ego history is all you need:

　　Rethinking the Open-Loop Evaluation of End-to-End Autonomous Driving in nuScenes　　

　　https://arxiv.org/abs/2305.10430

　　

 

 

2. Ego status  is all you need:

　　Is Ego Status All You Need for Open-Loop End-to-End Autonomous Driving?

　　https://arxiv.org/abs/2312.03031

　　这个文章和对应的知乎回答讲的特别好！

　　https://www.zhihu.com/question/350679956/answer/3308718224?utm_campaign=shareopn&utm_medium=social&utm_oi=861015326547132416&utm_psn=1731480516778536960&utm_source=wechat_session

 

 

3. Centerlines (& ego history) is all you need:

 

　　Parting with Misconceptions about Learning-based Vehicle Motion Planning

　　https://arxiv.org/abs/2306.07962

　　NUPLAN第一名的队伍总结的　　

　　https://www.zhihu.com/question/350679956/answer/3298582147