[KongchangAI]
Benchmark

MME-Emotion

A dataset for evaluating multimodal emotion recognition, containing 5,507 entries, used to measure model performance on emotion recognition tasks.

Timeline (last 90 days)

Nov 10

On 5,507 MME-Emotion samples, EmoHarness improved the task macro-average recognition rate of four student models and two Builders by 7.6 to 13.4 percentage points compared with single calls using video only.

Unverified50%

All Facts (1)

Source Articles