Summary of Citynav: Language-goal Aerial Navigation Dataset with Geographic Information, by Jungdae Lee et al.
CityNav: Language-Goal Aerial Navigation Dataset with Geographic Information
by Jungdae Lee, Taiki Miyanishi, Shuhei Kurita, Koya Sakamoto, Daichi Azuma, Yutaka Matsuo, Nakamasa Inoue
First submitted to arxiv on: 20 Jun 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
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 The paper presents a novel dataset called CityNav, designed for language-guided aerial navigation in photorealistic 3D environments of real cities. The dataset consists of 32k natural language descriptions paired with human demonstration trajectories, collected via a web-based 3D simulator. Each description identifies a navigation goal, utilizing the names and locations of landmarks within actual cities. The paper also proposes baseline models of navigation agents that incorporate an internal 2D spatial map representing landmarks referenced in the descriptions. The findings show that aerial agent models trained on human demonstration trajectories outperform those trained on shortest path trajectories by a large margin, and incorporating 2D spatial map information enhances navigation performance at a city scale. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper creates a new dataset called CityNav for guiding autonomous agents through real-world environments using visual and linguistic cues. The dataset has 32,000 descriptions of navigation goals that use landmarks in real cities. Researchers can use this data to train models that navigate through cities using language instructions. The results show that training models on human-made trajectories is better than training on shortest paths. Adding a spatial map to the models also improves their ability to navigate. |
