Real Grid RIR Dataset

This dataset contains real collected room impulse response (RIR) from a meeting room. The dataset is a location-specific RIR dataset where microphones are static and sound sources are distributed on a grid on the room floor. The RIR is collected using the sine sweep method and is collected using an eight-channel microphone.

This dataset has three subdatasets:

The first dataset was measured in all four directions at each of the 98 source locations (up, down, left, right). The microphone is placed on the center of the room.

The second dataset was measured only once at all 98 locations, with the microphone located in the center of the room.

The third dataset placed the microphone in the upper right corner of the room and measured the RIR once at each of the 98 locations.

This dataset is suitable for spatial-based research, especially distance-related, such as sound source distance estimation, distance based sound separation, and extraction tasks. The measurement environment and the room configuration are shown in the figure. The room size is about 5.9×6.9×2.9 meters with an RT60 of 0.60s.

The preprocessing process can be found at https://github.com/RunwuShi/GridRIRDataset?tab=readme-ov-file