MLLM framework detects hidden emotions with temporal context
The work builds a benchmark with expressed, internal emotion and concealment annotations to distinguish ordinary inconsistency from concealment using MLLM.
1 stories, newest first · one card per story
The work builds a benchmark with expressed, internal emotion and concealment annotations to distinguish ordinary inconsistency from concealment using MLLM.
Today 1.6K readers · 2.5K views · 779 reads · all time 189K views