Collecting accurate and precise emotion ground truth labels for mobile video watching is essential for ensuring meaningful predictions. However, video-based emotion annotation techniques either rely on post-stimulus discrete self-reports, or allow real-time, continuous emotion annotations (RCEA) only for desktop settings. Following a user-centric approach, we designed an RCEA technique for mobile video watching, and validated its usability and reliability in a controlled, indoor (N=12) and later outdoor (N=20) study. Drawing on physiological measures, interaction logs, and subjective workload reports, we show that (1) RCEA is perceived to be usable for annotating emotions while mobile video watching, without increasing users' mental workload (2) the resulting time-variant annotations are comparable with intended emotion attributes of the video stimuli (classification error for valence: 8.3%; arousal: 25%). We contribute a validated annotation technique and associated annotation fusion method, that is suitable for collecting fine-grained emotion annotations while users watch mobile videos.@en

Instead of predicting just one emotion for one activity (e.g., video watching), fine-grained emotion recognition enables more temporally precise recognition. Previous works on fine-grained emotion recognition require segment-by-segment, fine-grained emotion labels to train the recognition algorithm. However, experiments to collect these labels are costly and time-consuming compared with only collecting one emotion label after the user watched that stimulus (i.e., the post-stimuli emotion labels). To recognize emotions at a finer granularity level when trained with only post-stimuli labels, we propose an emotion recognition algorithm based on Deep Multiple Instance Learning (EDMIL) using physiological signals. EDMIL recognizes fine-grained valence and arousal (V-A) labels by identifying which instances represent the post-stimuli V-A annotated by users after watching the videos. Instead of fully-supervised training, the instances are weakly-supervised by the post-stimuli labels in the training stage. The V-A of instances are estimated by the instance gains, which indicate the probability of instances to predict the post-stimuli labels. We tested EDMIL on three different datasets, CASE, MERCA and CEAP-360VR, collected in three different environments: desktop, mobile and HMD-based Virtual Reality, respectively. Recognition results validated with the fine-grained V-A self-reports show that for subject-independent 3-class classification (high/neutral/low), EDMIL obtains promising recognition accuracies: 75.63% and 79.73% for V-A on CASE, 70.51% and 67.62% for V-A on MERCA and 65.04% and 67.05% for V-A on CEAP-360VR. Our ablation study shows that all components of EDMIL contribute to both the classification and regression tasks. Our experiments also show that (1) compared with fully-supervised learning, weakly-supervised learning can reduce the problem of overfitting caused by the temporal mismatch between fine-grained annotations and physiological signals, (2) instance segment lengths between 1-2 s result in the highest recognition accuracies and (3) EDMIL performs best if post-stimuli annotations consist of less than 30% or more than 60% of the entire video watching.@en