Virtual Reality (VR) video games have become popular in recent years with the emergence of a wide variety of enhanced VR hardware systems. Typically, current VR video games incorporate high-quality auditory and video feedback along with vibro-tactile cues to provide an immersive experience. However, the ongoing VR hardware and software do not consider the state of the user to determine whether the video game is generating an enjoyable experience, or if it has become a stressing or boring experience. This work provides an assessment of stress level estimation from features extracted from Electrocardiography (ECG), Electrodermal Activity (EDA), and Electromyography (EMG) signals of users while playing a VR video game with different difficulty levels. Statistical differences were found between the rest and gaming stages for several extracted features. Regardless of the fact that no significant statistical differences between the three levels of difficulty were found by analyzing the EDA and ECG features, an 83.1% accuracy was obtained for the classification between the three levels with a KNN model. For the EMG signal, the obtained accuracy ranged between 99% and 100% for the distinction between a difficult level 1 and rest stage for all models. When all features from the ECG, EDA and EMG signals were used, an accuracy of 99% was obtained for the differentiation between the three difficulty levels and a resting stage. The presented work results may serve as a reference for future work regarding feature extraction and game difficulty adaptation for user experience enhancement.

Virtual reality (VR) is continuing to grow as more affordable technological devices become available. Video games are one of the most profitable applications, while rehabilitation has the most significant social impact. Both applications require a proper user evaluation to provide personalized experiences that avoid boring or stressful situations. Despite the successful applications, there are several opportunities to improve the field of human–machine interactions, one of the most popular ones being the use of affect detection to create personalized experiences. In that sense, this study presents the implementation of two dynamic difficulty adaptation strategies. The person’s affective state is estimated through a machine learning classification model, which later serves to adapt the difficulty of the video game online. The results show that it is possible to maintain the user at a given difficulty level, which is analogous to achieving the well-known flow state. Among the two implemented strategies, no statistical differences were found in the workload induced by the users. However, more physical demands and a higher frustration were induced by one of the strategies, validated with the recorded muscular activity. The results obtained contribute to the state of the art of DDA strategies in virtual reality driven by affective data.