A review of current capabilities and future directions in machine-based emotion recognition

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not reported — no pooled effect size calculated

• Primary Outcome: No formally defined primary outcome. The review descriptively summarises technical approaches to machine-based emotion recognition across multiple modalities without synthesising a primary performance metric

• Nnt Or Sensitivity: No NNT, sensitivity, specificity, or diagnostic accuracy statistics are pooled or reported at the review level. Individual modality performance figures (e.g., classification accuracy for EEG-based or FER-based systems) may appear in cited primary studies but are not synthesised

• Confidence Interval: Not reported

Clinical Application

Feasibility in clinical settings is not assessed. Key barriers unaddressed include: regulatory approval requirements for medical devices, data privacy legislation compliance (e.g., Australian Privacy Act 1988, GDPR), integration with electronic health records, clinician training requirements, cost-effectiveness, and performance in diverse real-world clinical environments versus controlled laboratory conditions No Australian-specific content is present. Relevant Australian considerations include: TGA regulatory pathways for AI-based medical devices (Software as a Medical Device, SaMD framework), RACGP guidance on digital health tools in primary care, the Australian Digital Health Agency's national digital health strategy, and the significant cultural and linguistic diversity of the Australian population which directly impacts the generalisability of emotion recognition systems trained predominantly on Western, WEIRD (Western, Educated, Industrialised, Rich, Democratic) datasets. PBS listing of any related therapeutic technology is not applicable at this stage of evidence maturity. Mental health applications would need to align with Beyond Blue, headspace, and NDIS digital health frameworks before clinical deployment could be considered Potentially relevant to clinical populations where affective state monitoring may have therapeutic or diagnostic value — including patients with depression, anxiety disorders, autism spectrum disorder, dementia, chronic pain, and rehabilitation settings. However, the review does not validate any modality in a specific clinical population, and direct applicability to patient care cannot be established from this review alone

Abstract

The ability of machines to recognize emotions automatically is becoming increasingly significant across many domains where emotional understanding is essential. Such technology is applied in customer interaction, marketing, healthcare, education, the automotive industry, entertainment, and security. Providing real-time insights into human affective states improves user engagement and enables systems to respond more intelligently. Nevertheless, progress in this field is hindered by the inherent complexity of emotions, cultural differences in expression, and technical limitations that make accurate detection challenging. This paper delivers a broad review of contemporary approaches to emotion recognition. It highlights techniques based on facial expression analysis (FER), oculometrics (OM), microexpressions identification (MER), and speech analysis (SER). Further attention is given to methods involving body posture, gesture, and gait, as well as tactile interaction, text-based emotion recognition, and methods based on self-reporting. In addition, physiological signal-driven methods are discussed in depth, including respiration signals (RS), galvanic skin response (GSR), electroencephalography (EEG), electromyography (EMG), skin temperature (SKT), cardiac signals (ECG, PPG, HRV), and touch dynamics (TD) analysis. This comprehensive overview lays the foundation for advancing research on machine-based emotion recognition.

References

1. 1.Gasz, R., Bougriche, Z., Osuchowski, J., & Tomaszewski, M. (2026). A review of current capabilities and future directions in machine-based emotion recognition. Cognitive, Affective & Behavioral Neuroscience. https://doi.org/10.1016/j.imavis.2020.104043 [Note: DOI as supplied by source metadata — independent verification recommended due to journal-DOI mismatch identified during appraisal]