Findings

EmoMM: Benchmarking and Steering MLLM for Multimodal Emotion Recognition under Conflict and Missingness

AICA-Bench: Holistically Examining the Capabilities of VLMs in Affective Image Content Analysis