EgoLink Challenge, MM 2026
Developed SAGE, a multimodal agent system for linking and reasoning over egocentric interaction evidence.
Role: Lead contributor to SAGE
Recognition & Service
Competition results, academic honors, and contributions to the research community.
Challenge results, research contributions, and award certificates.
Developed SAGE, a multimodal agent system for linking and reasoning over egocentric interaction evidence.
Role: Lead contributor to SAGE
A multimodal affective-computing challenge focused on understanding emotional content in visual art.
Role: Team member
Improved Emotion-LLaMA with Conv-Attention for robust multimodal emotion recognition under noisy conditions.
Role: Core contributor
Used multimodal language-model reasoning to support open-vocabulary emotion recognition beyond a fixed label set.
Role: Core contributor
Built a multimodal system for extracting emotion-cause pairs from conversations using multimodal language models.
Role: Co-first author and core contributor
Introduced Expression MAE, a semi-supervised approach for learning robust multimodal emotion representations.
Role: First author and core contributor
Awarded first prize in the 13th China College Student Computer Design Competition.
Academic honors, scholarships, and talent development programs.
Selected for the doctoral-student track of the young science and technology talent development program.
University-level recognition awarded upon completion of the master's program.
Recognition for the quality and research contribution of the master's thesis.
National-level graduate scholarship recognizing academic and research excellence.
First-prize graduate academic scholarship, awarded in 2024.
First-prize graduate academic scholarship, awarded in 2023.
Scholarship received upon entering graduate study in 2022.
National Encouragement Scholarship, awarded in 2019.
Workshop and challenge organization, professional development, and peer review.
Selected for CAST's national studies and leadership development program for young science and technology professionals.
A workshop and challenge on neurophysiological intelligence for human-aware multimedia, co-located with MM 2026.
The MER2026 Challenge explores generative emotion understanding across interlocutor emotions, fine-grained labels, preferences, and physiological signals.
A VALSE 2026 Special Corner on organizing academic activities and bringing the research community together.
The MER2025 Challenge connects multimodal emotion recognition with large language models and generative emotion understanding.