Enhancing accuracy of 48 × 48 dimensional image using CNN

View at publisher Summarise with VIRAT
Authors: Trikalagga Saha, R. Gunavathi, Milan Chakraborty
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Machine learning (ML) is reshaping psychological research and practice by enabling data-driven, scalable analyses of complex variables, such as behaviour, cognition, and emotional responses, that traditional psychological models often cannot capture fully. ML methods allow for rapid processing of large datasets, offering insights into personality traits, mental health conditions, and behavioural patterns with a depth and adaptability beyond conventional approaches. ML methods facilitate rapid processing of extensive datasets, providing detailed insights into personality traits, mental health conditions, and behavioural patterns with adaptability and precision that extend beyond conventional methods. This paper presents a structured review of ML applications within psychology, with a focus on areas such as personality assessment, mental health diagnosis, behavioural prediction, and personalized therapy. It explores significant challenges, including model transparency, ethical issues, and the importance of cultural relevance, emphasizing both the transformative potential and limitations of ML in psychological contexts. Facial expression recognition plays a vital role in understanding human emotions, with applications in various domains. Building on the methodology proposed by Umer et al. [ 18 ], this study focuses on enhancing the accuracy of a generic convolutional neural network using improved data augmentation and a modified architecture. The model was trained and evaluated on the KDEF dataset, achieving a validation accuracy of 79.64%, which surpasses the benchmark accuracy reported in the base paper. These results demonstrate the effectiveness of the proposed approach in advancing the state of emotion recognition research.

Fields of Study & Subjects

Emotion and Mood Recognition Advanced Neural Network Applications Face recognition and analysis

Publication details

Journal / source
CRC Press eBooks
Pages
204–210
Publication date
25 March 2026
Language
EN
Document type
Book chapter