TUHIN ACHARJEE

Designed, developed, and implemented a rebar detection, tensile strength analysis, and robotic pick-and-place system, increasing testing efficiency by 60%. Programmed and optimized a robotic arm for precise pick-and-place operations of rebar, including real-time decision-making capabilities for streamlined testing processes. Leveraged advanced computer vision techniques and convolutional neural ne

Developed a video stabilization pipeline achieving 85% reduction in inter-frame jitter, using Lucas-Kanade optical flow and affine motion estimation across over 1,000 test frames. Tracked and filtered up to 200 keypoints per frame to estimate frame-wise transforms, then applied low-pass filtering on the accumulated camera trajectory for smooth temporal motion. Implemented geometric warping and fra

Built a custom thermal image dataset and trained a YOLOv8 model to detect high-temperature pipe leaks with ~90% [email protected]. Engineered a real-time inference pipeline with heatmap overlays and temperature-aware leak filtering for improved precision. Compared multiple CV models (YOLOv5, YOLOv8, UNet) through ablation studies to assess accuracy vs. speed trade-offs.

Grade: 8.5 CGPA

Completed a Bachelor of Technology in Computer Science and Engineering. Achieved a CGPA of 8.5.