arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.09622cs.CVcs.AIcs.LGcs.ROcs.SD
收藏

学习为视听导航设置航点

Learning to Set Waypoints for Audio-Visual Navigation

UT Austin(德克萨斯大学奥斯汀分校);Facebook AI Research(脸书人工智能研究院)

Changan Chen, Sagnik Majumder, Ziad Al-Halah, Ruohan Gao, Santhosh Kumar Ramakrishnan, Kristen Grauman

摘要:

在视听导航中,智能体使用视觉和声音智能地穿过复杂的、未映射的3D环境以找到声源(例如,另一个房间的电话响了)。现有的模型学习以固定的代理运动粒度进行操作,并依赖于音频观察的简单循环聚合。我们引入了一种强化学习方法,视听导航与两个关键的新元素:1)航点,动态设置和学习端到端的导航策略,和2)一个声学记忆,提供了一个结构化的,空间接地记录代理人听到了什么,因为它移动。这两个新的想法都利用了音频和视频数据的协同作用来揭示未映射空间的几何形状。我们在两个具有挑战性的真实世界3D场景数据集上展示了我们的方法,这两个数据集分别是CNET和Matterport3D。我们的模型大幅提高了最先进的技术水平,我们的实验表明,学习视觉,声音和空间之间的联系对于视听导航至关重要。项目:http://vision.cs.utexas.edu/projects/audio_visual_waypoints。

英文摘要:

In audio-visual navigation, an agent intelligently travels through a complex, unmapped 3D environment using both sights and sounds to find a sound source (e.g., a phone ringing in another room). Existing models learn to act at a fixed granularity of agent motion and rely on simple recurrent aggregations of the audio observations. We introduce a reinforcement learning approach to audio-visual navigation with two key novel elements: 1) waypoints that are dynamically set and learned end-to-end within the navigation policy, and 2) an acoustic memory that provides a structured, spatially grounded record of what the agent has heard as it moves. Both new ideas capitalize on the synergy of audio and visual data for revealing the geometry of an unmapped space. We demonstrate our approach on two challenging datasets of real-world 3D scenes, Replica and Matterport3D. Our model improves the state of the art by a substantial margin, and our experiments reveal that learning the links between sights, sounds, and space is essential for audio-visual navigation. Project: http://vision.cs.utexas.edu/projects/audio_visual_waypoints.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑