Pulkit Gautam

Summary

Computer Vision Engineer with 2+ years of experience in real-time systems, thermal imaging, and GPU-accelerated AI deployment. Strong background in deep learning, multisensor fusion, and image processing with hands-on experience in deploying models on edge platforms (NVIDIA Jetson). Experience spans thermal target detection, image super resolution, image stabilization, and real-time video pipelines.

Work Experience

Project Engineer — IRDE, DRDO, Dehradun

September 2024 – Present

Developed real-time computer vision pipelines for MWIR thermal imaging systems, including AI-based target detection, multisensor image fusion, and digital image stabilization
Designed and deployed GPU-based deep learning inference pipelines (Python) for real-time performance on edge platforms
Implemented thermal image processing and enhancement techniques, including non-uniformity correction (NUC) and Bad Pixel Replacement (BPR)
Built and optimized image processing modules for FPGA-based video pipelines (Verilog) for real-time thermal imaging systems
Integrated multiple sensors (thermal, visible, LRF, GPS) and performed system-level calibration, validation, and testing of electro-optical systems under real-world conditions
Developed sensor interfacing, communication protocols, and system interface documentation (EICD) for embedded electro-optical systems

Intern — IRDE, DRDO, Dehradun

January 2024 – June 2024

Built and deployed a real-time thermal target detection pipeline on NVIDIA Jetson AGX Orin
Created and annotated thermal imaging datasets for model training and evaluation
Implemented and benchmarked models like YOLOv7, Faster R-CNN using TensorFlow/PyTorch, analyzing latency vs accuracy trade-offs for real-time thermal detection

Projects

Deep-Learning Based Real-Time Target Detection in Thermal Imaging Systems

Implemented and benchmarked models like YOLO v7, RCNN and Faster R-CNN using TensorFlow/PyTorch real-time thermal detection on GPU based platform

Design and Simulation of Approximate Parallel Prefix Adder with Carry-Skip in Vivado

Designed an approximate prefix adder in Verilog and analyzed accuracy vs delay vs hardware trade-offs

Object Detection and Counting of Objects by Shape and Colour using Blob Analysis in MATLAB

Implemented using Blob Analysis Algorithm, Segmentation and Contour Detection in MATLAB

Certifications

The Joy of Computing using Python 2023 — Prof. Sudarshan Iyengar, IIT Ropar – NPTEL
An Introduction to Artificial Intelligence 2022 — Prof. Mausam, IIT Delhi – NPTEL
HTML, CSS, and JavaScript for Web Developers 2020 — Johns Hopkins University – Coursera

Computer Vision Engineer

Summary

Technical Skills

Work Experience

Projects

Education

Publications

Certifications