resource · 2026
Multimodal TCM Seven-Emotion Recognition: A Three-Modality Joint Consultation Framework Based on LLMs and Cross-Attention Fusion
proceedings-article
Multimodal TCM Seven-Emotion Recognition: A Three-Modality Joint Consultation Framework Based on LLMs and Cross-Attention Fusion is a resource in the TCM LLM catalog (Awesome-TCM-LLM), dated 2026. It is maintained alongside the GitHub list of Traditional Chinese Medicine large language models.
The seven-emotion framework reached a macro F1 of 87.6% on 10,593 IEMOCAP samples, then used a frozen large language model to turn the neural predictions into traditional Chinese medicine syndrome-differentiation reports.