Xiaomi has open-sourced MiDashengLM-Spatial, an audio-language model that combines general sound understanding with spatial awareness. The company released model weights and inference code under the Apache 2.0 license.
Built on Xiaomi’s MiDashengLM audio model, the system processes binaural audio to identify sound sources and reason about their positions and spatial relationships. It retains capabilities such as speech recognition, audio captioning and questions about environmental sounds and music.
Xiaomi plans to bring the technology to smart-home products and vehicle cabins, enabling devices to respond to both what they hear and where the sound originates. The release extends its Dasheng audio-model series beyond recognizing sound content to understanding the surrounding acoustic environment. [Xiaomi]