Photoplethysmography (PPG)-based continuous heart rate monitoring is essential in a number of domains, eg, for healthcare or fitness applications Recently, methods based on time-frequency spectra emerged to address the challenges of motion artefact compensation However, existing approaches are highly parametrised and optimised for specific scenarios of small, public datasets We address this fragmentation by contributing research into the robustness and generalisation capabilities of PPG-based heart rate estimation approaches First, we introduce a novel large-scale dataset (called PPG-DaLiA), including a wide range of activities performed under close to real-life conditions Second, we extend a state-of-the-art algorithm, significantly improving its performance on several datasets Third, we introduce deep learning to this domain, and investigate various convolutional neural network architectures Our end-to-end learning approach takes the time-frequency spectra of synchronised PPG- and accelerometer-signals as input, and provides the estimated heart rate as output Finally, we compare the novel deep learning approach to classical methods, performing evaluation on four public datasets We show that on large datasets the deep learning model significantly outperforms other methods: The mean absolute error could be reduced by 31 % on the new dataset PPG-DaLiA, and by 21 % on the dataset WESAD

https://www.mdpi.com/1424-8220/19/14/3079/pdf

Deep PPG: Large-Scale Heart Rate Estimation with Convolutional Neural Networks.

There is a rich repertoire of methods for stress detection using various physiological signals and algorithms. However, there is still a gap in research efforts moving from laboratory studies to real-world settings. A small number of research has verified when a physiological response is a reaction to an extrinsic stimulus of the participant’s environment in real-world settings. Typically, physiological signals are correlated with the spatial characteristics of the physical environment, supported by video records or interviews. The present research aims to bridge the gap between laboratory settings and real-world field studies by introducing a new algorithm that leverages the capabilities of wearable physiological sensors to detect moments of stress (MOS). We propose a rule-based algorithm based on galvanic skin response and skin temperature, combing empirical findings with expert knowledge to ensure transferability between laboratory settings and real-world field studies. To verify our algorithm, we carried out a laboratory experiment to create a “gold standard” of physiological responses to stressors. We validated the algorithm in real-world field studies using a mixed-method approach by spatially correlating the participant’s perceived stress, geo-located questionnaires, and the corresponding real-world situation from the video. Results show that the algorithm detects MOS with 84% accuracy, showing high correlations between measured (by wearable sensors), reported (by questionnaires and eDiary entries), and recorded (by video) stress events. The urban stressors that were identified in the real-world studies originate from traffic congestion, dangerous driving situations, and crowded areas such as tourist attractions. The presented research can enhance stress detection in real life and may thus foster a better understanding of circumstances that bring about physiological stress in humans.

https://www.mdpi.com/1424-8220/19/17/3805/pdf

Detecting Moments of Stress from Measurements of Wearable Physiological Sensors.

We exploit a self-supervised deep multi-task learning framework for electrocardiogram (ECG) -based emotion recognition. The proposed solution consists of two stages of learning a) learning ECG representations and b) learning to classify emotions. ECG representations are learned by a signal transformation recognition network. The network learns high-level abstract representations from unlabeled ECG data. Six different signal transformations are applied to the ECG signals, and transformation recognition is performed as pretext tasks. Training the model on pretext tasks helps the network learn spatiotemporal representations that generalize well across different datasets and different emotion categories. We transfer the weights of the self-supervised network to an emotion recognition network, where the convolutional layers are kept frozen and the dense layers are trained with labelled ECG data. We show that the proposed solution considerably improves the performance compared to a network trained using fully-supervised learning. New state-of-the-art results are set in classification of arousal, valence, affective states, and stress for the four utilized datasets. Extensive experiments are performed, providing interesting insights into the impact of using a multi-task self-supervised structure instead of a single-task model, as well as the optimum level of difficulty required for the pretext self-supervised tasks.

Self-supervised ECG Representation Learning for Emotion Recognition

The seminal work on Affective Computing in 1995 by Picard set the base for computing that relates to, arises from, or influences emotions. Affective computing is a multidisciplinary field of research spanning the areas of computer science, psychology, and cognitive science. Potential applications include automated driver assistance, healthcare, human-computer interaction, entertainment, marketing, teaching and many others. Thus, quickly, the field acquired high interest, with an enormous growth of the number of papers published on the topic since its inception. This paper aims to (1) Present an introduction to the field of affective computing though the description of key theoretical concepts; (2) Describe the current state-of-the-art of emotion recognition, tracing the developments that helped foster the growth of the field; and lastly, (3) point the literature take-home messages and conclusions, evidencing the main challenges and future opportunities that lie ahead, in particular for the development of novel machine learning (ML) algorithms in the context of emotion recognition using physiological signals.

/pdf/a-review-current-challenges-and-future-possibilities-on-365dzvqva1.pdf

A Review, Current Challenges, and Future Possibilities on Emotion Recognition Using Machine Learning and Physiological Signals

Affect recognition is an interdisciplinary research field bringing together researchers from natural and social sciences. Affect recognition research aims to detect the affective state of a person based on observables, with the goal to, for example, provide reasoning for the person's decision making or to support mental wellbeing (e.g., stress monitoring). Recently, beside of approaches based on audio, visual or text information, solutions relying on wearable sensors as observables, recording mainly physiological and inertial parameters, have received increasing attention. Wearable systems enable an ideal platform for long-term affect recognition applications due to their rich functionality and form factor, while providing valuable insights during everyday life through integrated sensors. However, existing literature surveys lack a comprehensive overview of state-of-the-art research in wearable-based affect recognition. Therefore, the aim of this paper is to provide a broad overview and in-depth understanding of the theoretical background, methods and best practices of wearable affect and stress recognition. Following a summary of different psychological models, we detail the influence of affective states on the human physiology and the sensors commonly employed to measure physiological changes. Then, we outline lab protocols eliciting affective states and provide guidelines for ground truth generation in field studies. We also describe the standard data processing chain and review common approaches related to the preprocessing, feature extraction and classification steps. By providing a comprehensive summary of the state-of-the-art and guidelines to various aspects, we would like to enable other researchers in the field to conduct and evaluate user studies and develop wearable systems.

https://www.mdpi.com/1424-8220/19/19/4079/pdf

Wearable-Based Affect Recognition-A Review.

Affect recognition aims to detect a person's affective state based on observables, with the goal to e.g. improve human-computer interaction. Long-term stress is known to have severe implications on wellbeing, which call for continuous and automated stress monitoring systems. However, the affective computing community lacks commonly used standard datasets for wearable stress detection which a) provide multimodal high-quality data, and b) include multiple affective states. Therefore, we introduce WESAD, a new publicly available dataset for wearable stress and affect detection. This multimodal dataset features physiological and motion data, recorded from both a wrist- and a chest-worn device, of 15 subjects during a lab study. The following sensor modalities are included: blood volume pulse, electrocardiogram, electrodermal activity, electromyogram, respiration, body temperature, and three-axis acceleration. Moreover, the dataset bridges the gap between previous lab studies on stress and emotions, by containing three different affective states (neutral, stress, amusement). In addition, self-reports of the subjects, which were obtained using several established questionnaires, are contained in the dataset. Furthermore, a benchmark is created on the dataset, using well-known features and standard machine learning methods. Considering the three-class classification problem ( baseline vs. stress vs. amusement ), we achieved classification accuracies of up to 80%,. In the binary case ( stress vs. non-stress ), accuracies of up to 93%, were reached. Finally, we provide a detailed analysis and comparison of the two device locations ( chest vs. wrist ) as well as the different sensor modalities.

Introducing WESAD, a Multimodal Dataset for Wearable Stress and Affect Detection

Affect recognition aims to detect a person's affective state based on observables, with the goal to e.g. provide reasoning for decision making or support mental wellbeing. Recently, besides approaches based on audio, visual or text information, solutions relying on wearable sensors as observables (recording mainly physiological and inertial parameters) have received increasing attention. Wearable systems offer an ideal platform for long-term affect recognition applications due to their rich functionality and form factor. However, existing literature lacks a comprehensive overview of state-of-the-art research in wearable-based affect recognition. Therefore, the aim of this paper is to provide a broad overview and in-depth understanding of the theoretical background, methods, and best practices of wearable affect and stress recognition. We summarise psychological models, and detail affect-related physiological changes and their measurement with wearables. We outline lab protocols eliciting affective states, and provide guidelines for ground truth generation in field studies. We also describe the standard data processing chain, and review common approaches to preprocessing, feature extraction, and classification. By providing a comprehensive summary of the state-of-the-art and guidelines to various aspects, we would like to enable other researchers in the field of affect recognition to conduct and evaluate user studies and develop wearable systems.

Wearable affect and stress recognition: A review.

Affective computing aims to detect a person's affective state (e.g. emotion) based on observables. The link between affective states and biophysical data, collected in lab settings, has been established successfully. However, the number of realistic studies targeting affect detection in the wild is still limited. In this paper we present an exploratory field study, using physiological data of 11 healthy subjects. We aim to classify arousal, State-Trait Anxiety Inventory (STAI), stress, and valence self-reports, utilizing feature-based and convolutional neural network (CNN) methods. In addition, we extend the CNNs to multi-task CNNs, classifying all labels of interest simultaneously. Comparing the F1 score averaged over the different tasks and classifiers the CNNs reach an 1.8% higher score than the classical methods. However, the F1 scores barely exceed 45%. In the light of these results, we discuss pitfalls and challenges for physiology-based affective computing in the wild.

Multi-target affect detection in the wild: an exploratory study

Multi-target affect detection in the wild: an exploratory study.

In affective computing (AC) field studies it is impossible to obtain an objective ground truth. Hence, self-reports in form of ecological momentary assessments (EMAs) are frequently used in lieu of ground truth. Based on four paradigms, we formulate practical guidelines to increase the accuracy of labels generated via EMAs. In addition, we detail how these guidelines were implemented in a recent AC field study of ours. During our field study, 1081 EMAs were collected from 10 subjects over a duration of 148 days. Based on these EMAs, we perform a qualitative analysis of the effectiveness of our proposed guidelines. Furthermore, we present insights and lessons learned from the field study.

Philip Schmidt

Papers

Introducing WESAD, a Multimodal Dataset for Wearable Stress and Affect Detection

Wearable affect and stress recognition: A review.

Multi-target affect detection in the wild: an exploratory study

Multi-target affect detection in the wild: an exploratory study.

Labelling Affective States "in the Wild": Practical Guidelines and Lessons Learned