• Developing gaze estimation models using Computer Vision and Deep Learning for Affective Computingin education, leveraging large-scale datasets.• Analyzing correlations between gaze patterns and student engagement in classroom settings to establish quantitative attention assessment measures.

Projects Completed:Physiological Signal Extraction Using Non-Invasive Techniques, Real-Time Emotion Detection using Gaze Features, Spatial 3D Curve Generation, Project Visio

Optimized a SAM-based image segmentation GUI, enabling efficient AI-assisted annotation, and developed an automated mechanism to aggregate segmented regions into sub-scenes, enhancing the semantic understanding of complex images.

Responsibilities:1. Explore and develop deep learning architectures tailored to the task of damage segmentation in wall paintings and their restoration.2. Utilize Convolutional Neural Networks (CNNs), known for their exceptional performance in image analysis tasks, as the foundation of these architectures.3. Investigate state-of-the-art CNN models, including U-Net, DeepLabv3+, PConvNet, AOT-GAN, and their variants, to identify the most suitable approaches.4. Adapt and customize the selected CNN models to address the unique characteristics and challenges presented by wall paintings.5. Focus on achieving accurate damage segmentation and effective restoration in the ancient murals.

Worked on the advancement of the college’s Mars Rover as part of various global inter-university rovercompetitions.