Summary of Activead: Planning-oriented Active Learning For End-to-end Autonomous Driving, by Han Lu et al.
ActiveAD: Planning-Oriented Active Learning for End-to-End Autonomous Driving
by Han Lu, Xiaosong Jia, Yichen Xie, Wenlong Liao, Xiaokang Yang, Junchi Yan
First submitted to arxiv on: 5 Mar 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Robotics (cs.RO)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper explores the problem of achieving sample and label efficiency for end-to-end autonomous driving (AD). The authors identify that one main bottleneck is the need for high-quality labeled data, which can be expensive to annotate. They propose a planning-oriented active learning method that progressively annotates raw data based on diversity and usefulness criteria for planning routes. This approach outperforms general active learning methods by a large margin and achieves comparable performance with state-of-the-art end-to-end AD methods using only 30% of the nuScenes dataset. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making self-driving cars work better using less data. Right now, it takes a lot of time and money to label all the images and videos needed for self-driving cars to learn. The authors found that most of this data is easy and not very useful, so they came up with a new way to pick the most important parts of the data and use them to train the car. This approach works really well and can even match the performance of more complex methods using much less data. |
Keywords
» Artificial intelligence » Active learning
