Skip to content

Dataset · HFUT-MMH-Lab

Multimodal dataset for mental disorder diagnosis

HFUT-MMH-Lab/multimodal_mental_health_dataset_MMH

Participant audio, video, and texts in English and Chinese with clinically confirmed diagnoses for distinguishing disorders.

What's inside

The dataset collects audio, video, and text recordings from each participant during five psychological tasks. Data is presented in English and Chinese, with a total volume ranging from 10 to 100 thousand records. All diagnostic labels are clinically confirmed by licensed specialists.

What you can do with it

  • Training multimodal models for differential diagnosis of mental disorders
  • Training text, speech, or video classifiers for psychological state analysis
  • Comparing and evaluating quality of emotion recognition and clinical feature algorithms
Tasks
Audio classification, Video classification, Text classification
Languages
English, Chinese
Size
10K–100K rows
Downloads this month
5.3K
Likes
5

How to load it

from datasets import load_dataset

ds = load_dataset("HFUT-MMH-Lab/multimodal_mental_health_dataset_MMH")