Speech emotion recognition (SER) has become increasingly important in areas such as healthcare, customer service, robotics, and human–computer interaction. The progress of this field depends not only on advances in algorithms but also on the databases that provide the training material for SER systems. These resources set the boundaries for how well models can generalize across speakers, contexts, and cultures. In this paper, we present a narrative review and comparative analysis of emotional speech corpora released up to mid-2025, bringing together both psychological and technical perspectives. Rather than following a systematic review protocol, our approach focuses on providing a critical synthesis of more than fifty corpora covering acted, elicited, and natural speech. We examine how these databases were collected, how emotions were annotated, their demographic diversity, and their ecological validity, while also acknowledging the limits of available documentation. Beyond description, we identify recurring strengths and weaknesses, highlight emerging gaps, and discuss recent usage patterns to offer researchers both a practical guide for dataset selection and a critical perspective on how corpus design continues to shape the development of robust and generalizable SER systems.

Background: Obesity remains a major global health challenge, and a significant proportion of bariatric surgery patients continue to experience dysfunctional emotional eating and body image concerns after surgery. Neurofeedback training (NFT) has been investigated as a potential intervention for maladaptive eating behaviours, but evidence in post-bariatric populations is still limited. Methods: Thirty-six patients who underwent sleeve gastrectomy were included, divided into an NFT group (N = 18) and a control group (N = 18). Assessments were performed at baseline and after 10 NFT sessions, using the Eating Disorder Inventory (EDI) and the Body Uneasiness Test (BUT). The intervention aimed to enhance alpha and theta waves with real-time feedback. Results: Compared with the control group, the NFT group showed significant improvements; specifically, reductions were observed in EDI subscales such as Drive for Thinness (p = 0.023, d = 0.51), Bulimia (p = 0.008, d = 0.92), Body Dissatisfaction (p = 0.015, d = 0.52), Ineffectiveness (p = 0.002, d = 0.89), Perfectionism (p = 0.006, d = 0.70), Interpersonal Distrust (p = 0.008, d = 0.82), and Interoceptive Awareness (p = 0.001, d = 0.91). Significant reductions were also found in BUT subscales including Weight Phobia (p = 0.041, d = 0.84), Body Image Concern (p = 0.039, d = 0.90), Avoidance (p = 0.027, d = 0.83), Compulsive Self-Monitoring (p = 0.013, d = 0.83), and Depersonalisation (p = 0.033, d = 0.85). Conclusions: The data indicate that NFT may help reduce emotional eating and related psychological factors in post-bariatric patients in the short term. However, studies with larger samples and longer follow-ups are needed to confirm its effectiveness and assess its clinical applicability.