HARISSH A S

• Conducted research on brain tumor detection using medical imaging, applying deep learning models YOLOv8, UNet, and ResNet-50 for tumor classification and segmentation. • Led supervised image processing for over 900 MRI scans, utilizing contrast enhancement, histogram equalization, and data augmentation to optimize input quality for machine learning. • Produced analytical reports and obtained clinical validation from radiologists to initiate development of a deep learning model for early-stage diagnosis by 30%. • Trained and evaluated multiple architectures on public datasets (e.g., BraTS, Figshare), achieving a peak accuracy of 92.4%, Dice coefficient of 0.87, and IoU score of 0.81. • Benchmarked performance across traditional ML classifiers (SVM, Random Forest) using SIFT + BoVW features vs. deep CNNs, demonstrating a 15% improvement in F1-score with deep learning

• Promoted to Senior Intern for exceptional performance in ML research & deployment. • Led and mentored junior interns from Anna University across projects including pneumonia detection, retinal disease detection, myocardial scar identification, and satellite image crop classification. • Designed and deployed a Cattle Identification System using a lightweight YOLO-NAS model for muzzle detection, storing embeddings in FAISS/Quadrant DBs, and achieving 83–90% accuracy in under 3 weeks. • Automated processes with FastAPI services, improving scalability and real-world usability. • Enhanced the Facial Recognition Attendance System (FRAS) by building preprocessing pipelines and improving model robustness. • Developed and optimized Crop & Pest Detection models, localizing pests (Stemborer, Fall Armyworm, Leaffolder) and applying Segment Anything Model (SAM) for auto-annotations across 80 lakh images spanning 23 crops, achieving ~70% accuracy on real-time farmer-collected images. • Represented TNeGA in NVIDIA & AWS AI conferences, engaged with government departments (WRD, GIS), and interacted with AI/ML startups, delivering detailed reports to senior officials. • Volunteered at the TNeGA GIS Conference (attended by IAS, IPS, and Govt. officers), gaining exposure to high-level AI/ML applications in governance.

Coursework: Spring 2026: Advanced Deep Learning Embedded AI Large Scale Stream Processing Mathematics for Machine learning ,controls and signals Fall 2025: Big Data Analytics Reinforcement Learning Digital Signal Processing Continues Control System

Coursework: Fall 2025: Big Data Analytics Convex Optimization Digital Signal Processing Data Centre Processing