Sanjay Swami

• Developed and deployed deep learning models for onboard cloud segmentation and MWIR-based object detection • Designed and Docker-based deployment of end-to-end onboard data processing pipelines for resource-constrained environments • Defined concept of operations, flight operation procedures, and PDR/CDR documentation • Contributed to payload testing and validation (including vibration, TVAC, EMC) • Coordinated with cross-functional teams to support system integration and technical alignment

• Developed deep learning-based solutions for 6D object pose estimation in robotic applications, enabling accurate pose prediction from RGB-D data • Generated synthetic training datasets using the NDDS plugin in Unreal Engine, and designed pipelines for real-world data acquisition with ground truth pose annotation and segmentation • Trained and optimized deep learning models, including Pixel-wise Voting Network (PVNet) and DenseFusion, for robust 6D pose estimation • Evaluated model performance

Control Engineering, Robotics, Artificial Intelligence, Automation, and Dynamic and Mechatronic Systems Research project in image processing and computer vision using Deep Learning and OpenCV libraries

Finolex Academy of Management & Technology