AUTHOR=Alsubai Shtwai , Alqahtani Abdullah , Alanazi Abed , Sha Mohemmed , Gumaei Abdu 

TITLE=Facial emotion recognition using deep quantum and advanced transfer learning mechanism

JOURNAL=Frontiers in Computational Neuroscience

VOLUME=Volume 18 - 2024

YEAR=2024

URL=https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2024.1435956

DOI=10.3389/fncom.2024.1435956

ISSN=1662-5188

ABSTRACT=Facial expressions have become a common way for interaction among humans. People cannot comprehend and predict the emotions or expressions of individuals through simple vision. Thus, in psychology, detecting facial expressions or emotion analysis demand an assessment and evaluation of decisions for identifying the emotions of a person or any group during communication. With recent evolvement of technology, AI (Artificial Intelligence) have gained significant usage wherein DL based algorithms finds employability for detecting facial expressions. Considering this, the study proposes a system design that could detect facial expressions by extracting relevant features using the Modified ResNet model. Unlike existing works, proposed system stacks building-blocks with residual connections and advanced extraction method with quantum-computing that could accomplish computations with second for which recent methodologies would require more processing time. This is referred as quantum superiority that applies methods for extracting local and global features. For accomplishing feature extraction, backbone stem utilizes quantum convolutional layer that is encompassed of several parameterized quantum-filters. The research makes use of residual connections for joining the filter-connection phases in ResNet-18 model with the Modified up Sampled Bottle Neck Process (MuS-BNP). This will permit the model for retaining its computational efficacy while attaining the merits of residual connections. The proposed model is capable of overcoming the issues of maximum similarity within varied expressions. The performance of the overall proposed system is assessed with regard to performance metrics comprising of accuracy, F1-score, recall and precision. This performance analysis process will assist in confirming the efficacy of the proposed system.