Recognition of emotions from physiological signals, and in particular from electroencephalography (EEG), is a field within affective computing gaining increasing relevance. Although researchers have used these signals to recognize emotions, most of them only identify a limited set of emotional states (e.g., happiness, sadness, anger, etc.) and have not attempted to predict exact values for valence and arousal, which would provide a wider range of emotional states. This paper describes our proposed model for predicting the exact values of valence and arousal in a subject-independent scenario. To create it, we studied the best features, brain waves, and machine learning models that are currently in use for emotion classification. This systematic analysis revealed that the best prediction model uses a KNN regressor (K = 1) with Manhattan distance, features from the alpha, beta and gamma bands, and the differential asymmetry from the alpha band. Results, using the DEAP, AMIGOS and DREAMER datasets, show that our model can predict valence and arousal values with a low error (MAE < 0.06, RMSE < 0.16) and a strong correlation between predicted and expected values (PCC > 0.80), and can identify four emotional classes with an accuracy of 84.4%. The findings of this work show that the features, brain waves and machine learning models, typically used in emotion classification tasks, can be used in more challenging situations, such as the prediction of exact values for valence and arousal.

Recognizing emotions from physiological signals has proven to be important in various scenarios. To assist in developing emotion recognizers, software frameworks and toolboxes have emerged, offering ready-to-use components. However,these have limitations regarding the type of physiological signals supported, the recognition steps covered, or the acquisition of multiple physiological signals. This paper presents metaFERA, an architectural meta-framework for creating software frameworks for end-to-end emotion recognition from physiological signals. The modularity and flexibility of the meta-framework and the resulting frameworks allow the fast prototyping of emotion recognition systems and experiments to test and validate new algorithms. To that end, metaFERA offers: (i) a set of pre-configured blocks to which we can add behavior to create framework components; (ii) an easy way to add behavior to the pre-configured blocks; (iii) a channel-based communication mechanism that transparently and efficiently supports the exchange of information between components; (iv) a simple and easy way to use and link components from a resulting framework to create applications. Additionally, we provide a set of Web services, already configured, to make the resulting recognition systems available as a service. To validate metaFERA, we created a framework for Electrodermal Activity, an emotion recognizer to identify high/low arousal using the aforementioned framework, and a layer to offer the recognizer as a service.

Caregivers are one of the main pillars in adopting digital technologies for reminiscence therapy, as they are responsible for its administration. Despite their comprehensive understanding of the needs and difficulties associated with the therapy, their perspective has not been fully considered in existing technological solutions. To mitigate this gap, we followed a user-centered design approach, using a sequential process that included worldwide web-based surveys, follow-up interviews, and focus groups to inform the development of novel assistive technological solutions focused on the real needs of caregivers. A total of 713 informal and 67 formal caregivers participated in our study. Our findings reveal that caregivers are overburdened with the amount of work they have daily; thus, mechanisms that could help them manage all the tasks involved in therapy (e.g., creating sessions, gathering feedback for further consultation) would help to reduce their workload while potentially improving the quality of therapy sessions. Caregivers also want to be able to easily leverage emotions during therapy to personalize and diversify sessions over time, thus preventing aggression or agitation in people with dementia. As a result of our study, we propose a list of functional requirements gathered for both formal and informal caregivers and the corresponding expected primary and secondary outcomes, such as improvement of the cognitive function of people with dementia, reduction of caregivers' stress and burden, and reduction of behavioral symptoms of people with dementia. We also present the resulting architecture of the KeepsakeBox prototype, which allows creating, managing, and delivering personalized reminiscence therapy sessions to people with dementia, which can be used as a basis for the development of future technological solutions for reminiscence therapy.

The success of supervised models for emotion recognition on images heavily depends on the availability of images properly annotated. Although millions of images are presently available, only a few are annotated with reliable emotional information. Current emotion recognition solutions either use large amounts of weakly-labeled web images, which often contain noise that is unrelated to the emotions of the image, or transfer learning, which usually results in performance losses. Thus, it would be desirable to know which images would be useful to be annotated to avoid an extensive annotation effort. In this paper, we propose a novel approach based on active learning to choose which images are more relevant to be annotated. Our approach dynamically combines multiple active learning strategies and learns the best ones (without prior knowledge of the best ones). Experiments using nine benchmark datasets revealed that: (i) active learning allows to reduce the annotation effort, while reaching or surpassing the performance of a supervised baseline with as little as 3% to 18% of the baseline's training set, in classification tasks; (ii) our online combination of multiple strategies converges to the performance of the best individual strategies, while avoiding the experimentation overhead needed to identify them.