用脑电信息指导面部情绪识别，提升跨数据集泛化能力。
NeuroGaze-Distill: Brain-informed Distillation and Depression-Inspired Geometric Priors for Robust Facial Emotion Recognition
- 通过脑电图启发的静态情绪原型和抑郁研究几何先验，实现跨模态知识蒸馏。
- 在FERPlus上达到89.2%准确率，跨数据集测试中显著优于基线模型。
- 无需部署时使用脑信号，适合实际应用中的鲁棒情绪识别场景。
仅基于像素训练的面部情绪识别（FER）模型在不同数据集间泛化能力差，因面部外观是情绪的间接且有偏的代理。本文提出NeuroGaze-Distill，一种跨模态知识蒸馏框架，将脑电图（EEG）信息中的先验知识注入仅依赖图像的FER学生模型。教师模型基于DREAMER数据集的头皮地形图训练，并以MAHNOB-HCI作为无标签支持，生成冻结的5×5价值/唤醒（V/A）原型网格；部署时无需脑信号或非视觉输入。学生模型（ResNet-18/50）在FERPlus上使用标准交叉熵与知识蒸馏，并引入两个轻量正则化项：(i) Proto-KD（余弦相似度）使学生特征对齐静态原型；(ii) D-Geo 软性调整嵌入空间几何结构，符合抑郁研究中的典型发现（如高价值区域出现类似快感缺失的收缩）。在域内（FERPlus验证集）和跨数据集协议（AffectNet-mini；可选CK+）下评估，报告标准8分类准确率及仅现有时的宏平均F1与平衡准确率，以公平处理标签集不匹配问题。消融实验表明原型和D-Geo均带来稳定提升，且5×5网格优于更密集的网格，兼顾稳定性。该方法简单、可部署，无需复杂架构即可增强鲁棒性。
原文摘要 · Abstract (English)
Facial emotion recognition (FER) models trained only on pixels often fail to generalize across datasets because facial appearance is an indirect and biased proxy for underlying affect. We present NeuroGaze-Distill, a cross-modal distillation framework that transfers brain-informed priors into an image-only FER student via static Valence/Arousal (V/A) prototypes and a depression-inspired geometric prior (D-Geo). A teacher trained on EEG topographic maps from DREAMER (with MAHNOB-HCI as unlabeled support) produces a consolidated 5x5 V/A prototype grid that is frozen and reused; no EEG-face pairing and no non-visual signals at deployment are required. The student (ResNet-18/50) is trained on FERPlus with conventional CE/KD and two lightweight regularizers: (i) Proto-KD (cosine) aligns student features to the static prototypes; (ii) D-Geo softly shapes the embedding geometry in line with affective findings often reported in depression research (e.g., anhedonia-like contraction in high-valence regions). We evaluate both within-domain (FERPlus validation) and cross-dataset protocols (AffectNet-mini; optional CK+), reporting standard 8-way scores alongside present-only Macro-F1 and balanced accuracy to fairly handle label-set mismatch. Ablations attribute consistent gains to prototypes and D-Geo, and favor 5x5 over denser grids for stability. The method is simple, deployable, and improves robustness without architectural complexity.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。