The paper introduces the SOUND CAM dataset, the largest collection of unique Room Impulse Responses (RIRs) from real-world rooms. The dataset includes 5,000 10-channel real-world measurements of room impulse responses and 2,000 10-channel recordings of music in three different types of rooms with varying human positions. The dataset aims to explore the use of a room’s acoustic properties for detecting and identifying humans and tracking their positions. The SOUND CAM dataset focuses on the effects of changing the positions and identities of humans on the acoustic properties of the room.
Publication date: 6 Nov 2023
Project Page: https://masonlwang.com/soundcam/
Paper: https://arxiv.org/pdf/2311.03517
