arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2604.05926cs.HC
收藏

FEEL:量化生理信号异质性以实现通用情感识别

FEEL: Quantifying Heterogeneity in Physiological Signals for Generalizable Emotion Recognition

IIIT-Delhi(德里信息与技术研究所);RIT(罗切斯特理工学院)

Pragya Singh, Ankush Gupta, Somay Jalan, Mohan Kumar, Pushpendra Singh

AI总结:

FEEL通过评估19个数据集上的EDA和PPG信号,探讨了情感识别中异质性的影响,发现CLSP模型在F1分数上表现最佳,而传统模型和手工特征方法也具有竞争力。

摘要:

情感识别从生理信号在心理健康和情感感知系统中有巨大潜力。然而,缺乏跨异质数据集的标准化大规模评估限制了进展和模型泛化。我们介绍了FEEL,这是首个大规模基准研究,评估了16种架构在19个公开数据集上的情感识别表现,包括不同实验设置、设备类型和标注策略的泛化分析。结果显示,微调的对比信号-语言预训练(CLSP)模型在 arousal 和 valence 分类任务中取得最高 F1 分数,而简单模型如随机森林、LDA 和 MLP 仍具竞争力。利用手工特征的模型在低资源、噪声环境下表现更优。进一步的跨数据集分析显示,训练于真实生活场景的数据在实验室(F1 = 0.79)和约束设置(F1 = 0.78)中泛化良好。同样,训练于专家标注数据的模型可有效转移到刺激标注(F1 = 0.72)和自我报告数据集(F1 = 0.76)。此外,训练于实验室设备的模型在自定义可穿戴设备(F1 = 0.81)和Empatica E4(F1 = 0.73)中也表现出高泛化能力,凸显了异质性的影响。更多关于FEEL的信息请访问我们的网站 https://alchemy18.github.io/FEEL_Benchmark/。

英文摘要:

Emotion recognition from physiological signals has substantial potential for applications in mental health and emotion-aware systems. However, the lack of standardized, large-scale evaluations across heterogeneous datasets limits progress and model generalization. We introduce FEEL, the first large-scale benchmarking study of emotion recognition using electrodermal activity (EDA) and photoplethysmography (PPG) signals across 19 publicly available datasets. We evaluate 16 architectures spanning traditional machine learning, deep learning, and self-supervised pretraining approaches, structured into four representative modeling paradigms. Our study includes both within-dataset and cross-dataset evaluations, analyzing generalization across variations in experimental settings, device types, and labeling strategies. Our results showed that fine-tuned contrastive signal-language pretraining (CLSP) models (71/114) achieve the highest F1 across arousal and valence classification tasks, while simpler models like Random Forests, LDA, and MLP remain competitive (36/114). Models leveraging handcrafted features (107/114) consistently outperform those trained on raw signal segments, underscoring the value of domain knowledge in low-resource, noisy settings. Further cross-dataset analyses reveal that models trained on real-life setting data generalize well to lab (F1 = 0.79) and constraint-based settings (F1 = 0.78). Similarly, models trained on expert-annotated data transfer effectively to stimulus-labeled (F1 = 0.72) and self-reported datasets (F1 = 0.76). Moreover, models trained on lab-based devices also demonstrated high transferability to both custom wearable devices (F1 = 0.81) and the Empatica E4 (F1 = 0.73), underscoring the influence of heterogeneity. More information about FEEL can be found on our website https://alchemy18.github.io/FEEL_Benchmark/.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑