NeuroMM Grand Challenge

Epilepsy remains a high-impact testbed for trustworthy AI, where Interictal Epileptiform Discharges (IEDs) are critical biomarkers for diagnosis and localization. Existing methods are often EEG-only and do not fully exploit multimodal context, leading to limited robustness under class imbalance, biological variability, and real-world noise.

NeuroMM 2026 addresses this gap by introducing a multimodal challenge framework that integrates physiological signals with synchronized contextual information. The challenge is designed to advance robust multimodal perception, cross-modal reasoning, and clinically actionable inference.

The challenge includes three official tracks: NMM-Basic-IED, NMM-Context-IED, and NMM-Source-IED, covering detection, context-aware robustness, and spatial localization.

• We encourage all participating teams to prepare their challenge solutions as papers and submit them to the NeuroMM Workshop 2026.
• The top three teams in each track will receive certificates, provided that they submit a paper and their code passes the official review.
• Three outstanding papers will be selected from all submitted challenge papers and recommended to the ACM MM main conference proceedings.
• Award evaluation will consider both the single-track ranking and the quality of the submitted paper.
• Each participating team can have at most one paper recommended to the ACM MM main conference proceedings.

Challenge Timeline

NeuroMM-2026 Challenge will officially start at 00:00 on May 21, 2026 (Beijing Time) and close at 24:00 on Jun 11, 2026 (Beijing Time). Please arrange your work carefully and complete submissions on time.

Code

The official baseline code and starter toolkit are now available on GitHub.

Advisory Committee

Fei Ma

Guangming Laboratory

Zitong Yu

Great Bay University

Zheng Lian

Tongji University

Laizhong Cui

Shenzhen University

Qi Tian

Guangming Laboratory & Huawei