Synheart Emotion: Privacy-Preserving On-Device Emotion Recognition from Biosignals
It addresses privacy concerns and real-time needs for wearable devices by enabling on-device emotion recognition, though it is incremental as it focuses on optimizing existing methods for a specific application.
This paper tackles the problem of privacy and latency in emotion recognition by evaluating on-device models using wrist-based biosignals, finding that classical ensemble methods like ExtraTrees outperform deep learning with an F1 score of 0.623 on wrist-only features and achieve a 4.08 MB footprint with 0.05 ms inference latency.
Human-computer interaction increasingly demands systems that recognize not only explicit user inputs but also implicit emotional states. While substantial progress has been made in affective computing, most emotion recognition systems rely on cloud-based inference, introducing privacy vulnerabilities and latency constraints unsuitable for real-time applications. This work presents a comprehensive evaluation of machine learning architectures for on-device emotion recognition from wrist-based photoplethysmography (PPG), systematically comparing different models spanning classical ensemble methods, deep neural networks, and transformers on the WESAD stress detection dataset. Results demonstrate that classical ensemble methods substantially outperform deep learning on small physiological datasets, with ExtraTrees achieving F1 = 0.826 on combined features and F1 = 0.623 on wrist-only features, compared to transformers achieving only F1 = 0.509-0.577. We deploy the wrist-only ExtraTrees model optimized via ONNX conversion, achieving a 4.08 MB footprint, 0.05 ms inference latency, and 152x speedup over the original implementation. Furthermore, ONNX optimization yields a 30.5% average storage reduction and 40.1x inference speedup, highlighting the feasibility of privacy-preserving on-device emotion recognition for real-world wearables.