Understanding Art Deeply: Sentiment Analysis of Facial Expressions of Graphic Arts Using Deep Learning

Art serves as a profound medium for humans to express and present their thoughts, emotions, and experiences in aesthetically and captivating means. It is like a universal language transcending the limitations of language enabling communication of complex ideas and feelings. Artificial Intelligence (AI) based data analytics are being applied for research domains such as sentiment analysis in which usually text data is analyzed for opinion mining. In this research study, we take art work and apply deep learning (DL) algorithms to classify seven diverse facial expressions in graphics art. For empirical analysis, state of the art deep learning algorithms of Inceptionv3 and pre-trained model of ResNet have been applied on large dataset. Both models are considered revolutionary deep learning architecture allowing for the training of much deeper networks and thus enhancing model performance in various computer vision tasks such as image recognition and classification tasks. The comprehensive results analysis reveals that the proposed methods of ResNet and Inceptionv3 have achieved accuracy as high as 98% and 99% respectively as compared to existing approaches in the relevant field. This research contributes to the fields of sentiment analysis, computational visual art, and human-computer interaction by addressing the detection of seven diverse facial expressions in graphic art. Our approach enables enhanced understanding of user sentiments, offering significant implications for improving user engagement, emotional intelligence in AI-driven systems, and personalized experiences in digital platforms. This study bridges the gap between visual aesthetics and sentiment detection, providing novel insights into how graphic art influences and reflects human emotions by highlighting the efficacy of DL frameworks for real-time emotion detection applications in diverse fields such as human psychological assessment and behavior analysis.

@article{Wang2025,
  title     = {Understanding Art Deeply: Sentiment Analysis of Facial Expressions of Graphic Arts Using Deep Learning},
  journal   = {International Journal of Advanced Computer Science and Applications},
  volume    = {16},
  number    = {1},
  year      = {2025},
  publisher = {The Science and Information Organization},
  author    = {Fei Wang},
  doi       = {10.14569/IJACSA.2025.0160152},
  url       = {https://doi.org/10.14569/IJACSA.2025.0160152}
}