CASE：面向冻结多模态情绪模型的技能演化框架
CASE: Confusion-Aware Skill Evolution for Frozen Multimodal Emotion Models
arXiv 上新论文 CASE，专门治多模态模型分不清相近情绪的问题，把 Qwen3-VL-8B 的 CA-MER 准确率从 46.80% 拉到 59.13%，还不用动模型参数。
CASE 是首个面向多模态情绪识别（MER）的 agent-skill 框架，由 Confusion-Aware Skill Evolution 团队提出。它维护一个稀疏、可自我进化的 EmotionSkills 库，针对语义相近情绪对存储专门的成对技能，推理时动态检索以增强边界感知能力，且无需改动模型参数。在 CA-MER、DFEW 和 EMER 三个基准上的实验验证了其跨模型迁移性。CASE 将 Qwen3-VL-8B 在 CA-MER 基准上的平均准确率从 46.80% 提升至 59.13%，并能迁移到多种冻结的 MLLM 骨干模型。
CASE: Confusion-Aware Skill Evolution for Frozen Multimodal Emotion Models
Multimodal large language models (MLLMs) have substantially advanced Multimodal Emotion Recognition (MER) by enabling fine-grained emotion understanding through reasoning over multimodal evidence. However, they still exhibit systematic confusion between semantically adjacent emotions with highly overlapping multimodal cues. Existing approaches typically address these errors through global parameter adaptation, making confusion correction costly and inflexible. We instead observe that emotion confusion is highly localized and therefore better addressed through reusable pair-specific expertise. Building on this insight, we propose Confusion-Aware Skill Evolution (CASE), the first agent-skill framework for MER. CASE maintains a sparse, self-evolving library of pair-specific EmotionSkills that externalize affective expertise from model parameters, enabling the learned expertise to be continuously refined and readily transferred across diverse frozen MLLM backbones. During inference, CASE dynamically retrieves the appropriate EmotionSkill to reinforce boundary-aware reasoning while keeping the MLLM backbone frozen. Extensive experiments on CA-MER, DFEW, and EMER demonstrate the effectiveness and transferability of CASE across both general-purpose and emotion-specialized MLLMs. Notably, CASE improves the average accuracy of Qwen3-VL-8B on the CA-MER benchmark from 46.80% to 59.13%, while achieving consistent gains across MLLMs.