Welcome

Welcome to the home of the CHiME (Computational Hearing in Multisource Environments) challenges and corresponding workshops on Speech Processing in Everyday Environments.

Current

These challenges are currently running, with submissions due February 7, 2026.

CHiME-9 Task 1, MCoRec: Multi-Modal Context-aware Recognition

MCoRec targets the problem of Multi-Modal Context-aware Recognition in a single-room environment. The goal is to process a single 360° video and audio recording of a room where multiple, separate conversations are happening simultaneously, and to both transcribe each speaker’s speech and identify which speakers belong to the same conversation.

Get the dataset on the MCoRec Hugging Face page.

CHiME-9 Task 2, ECHI: Enhancing Conversations to address Hearing Impairment

ECHI invites participants to tackle the challenge of Enhancing Conversations to address Hearing Impairment. We consider signal processing that enables a device, which will block out unwanted noise but while still allowing the wearer to hear their conversational partners.

Get the dataset on the ECHI Hugging Face page.