Skip to main content
Darshi Dwivedi

Darshi Dwivedi

Computer Vision / Robotics engineer

Talent
AhmedabadMember since October 2, 2026
ROS2MoveItMoveIt2YOLOv8OpenCvTensorFlowPytorchDINOOpenVINOONNXSCADA
·

Hire this person

Send a job offer directly to this candidate

ROS2MoveItMoveIt2YOLOv8OpenCv

About

Computer Vision / Robotics engineer with hands-on experience building end-to-end CV pipelines — data collection, annotation, augmentation, training, and evaluation — using PyTorch, TensorFlow, and OpenCV. Experienced in optimizing and deploying models (ONNX, OpenVINO) for real-time inference on edge hardware, with additional work in 3D vision, multi-object tracking, and multi-camera fusion.

Experience

Manipulation Engineer at Bharat Forge (2025-07 – Present)

  • Designed, trained, and fine-tuned an end-to-end YOLOv8 object detection pipeline — dataset collection, annotation, augmentation, and model optimization — for multi-class pallet detection, achieving 96.3% mAP@50 on the validation dataset.
  • Optimized the trained model for latency and resource efficiency by converting it to ONNX and OpenVINO IR formats, enabling real-time inference deployment on Intel NUC edge hardware.
  • Built a vision-based perception pipeline using DINO (ViT-based) for object detection and 6D pose estimation, integrating outputs with MoveIt2 for downstream motion planning and autonomous pick-and-place execution.
  • Developed a 3D vision pipeline fusing 2D detections with RGB-D depth images and point clouds to estimate object pose (x, y, z, yaw).
  • Built a real-time multi-object tracking system using multi-camera and camera–LiDAR fusion for target detection, tracking, and autonomous guidance up to a 15m range.

Project Engineer at Encon Systems (2023-06 – 2025-03)

  • Built and deployed a TensorFlow & OpenCV-based object detection model on Raspberry Pi edge hardware, integrating it into robotic workflows to trigger classification-based alerts.
  • Developed a ROS2-based smart palletizing system with MoveIt motion planning, improving stacking precision and reducing cycle time.
  • Designed pallet pattern generation logic for automated box placement.
  • Developed a SCADA system for real-time machine monitoring and PLC communication.

Node.js Intern at Celebal Technologies (2022-06 – 2022-07)

  • Worked on backend development tasks and gained hands-on experience in real-world project workflows.

Education

B.Tech in Electronics & Communication Engineering – Arya College of Engineering & IT, Jaipur (2019 – 2023)

Class XII – Dungarpur Public School (2017 – 2018)

Class X – Kendriya Vidyalaya Dungarpur (2015 – 2016)

Skills

Computer Vision / Robotics engineer

Reviews

Related