Saurabh Kumar Singh

Analyze complex images and video for object detection, spatial relationships, visual inconsistencies, depth, geometry, and reasoning errors. Evaluate Vision-Language Model predictions, identify hallucinations and reasoning failures, and create corrective annotations and high-quality multimodal training and evaluation datasets.

Taught cybersecurity, cryptography, image processing, and C programming across undergraduate and postgraduate programs. Designed laboratory manuals and outcome-based assessments, mentored students on AI capstone projects, and coordinated academic timetables for approximately 35 faculty members.

Built a generative AI textile design system for creating customized fashion prints and visual patterns from user-defined inputs. Implemented image-generation, preprocessing, enhancement, and style-control workflows, and collaborated with product stakeholders to improve output quality, controllability, and consistency.