Kishore Kulanthaivel

Developed and deployed AI-powered industrial inspection systems using deep learning and real-time image processing. Built and fine-tuned YOLO, U-Net and SAM2 models for defect detection and segmentation. Developed end-to-end ML pipelines from data annotation and augmentation to training, evaluation and deployment. Worked with Python, PyTorch, TensorFlow/Keras, OpenCV, TensorRT, C++ and PyQt for real-time inspection, visualization and hardware integration.

• Gained practical experience in embedded systems, image processing, and deep learning edge integration. • Developed a proof-of-concept inspection system for the print industry using computer vision and AI for automated defect detection. • Built and optimized deep learning models to identify printing defects such as color variation, misalignment, and text inconsistencies. • Designed PyQt-based interfaces for real-time defect visualization, reporting, and workflow automation to improve production reliability