RAKTIMJYOTI PARASHAR

Philadelphia, PA P: +267-***-**** ad1mqo@r.postjobfree.com ad1mqo@r.postjobfree.com EDUCATION

UNIVERSITY OF PENNSYLVANIA Philadelphia, PA

Master of Science in Mechanical Engineering and Applied Mechanics May 2023 Concentration: Robotics and Mechatronics

Cumulative GPA: 3.68/4.0

Relevant Coursework: Machine Learning, Deep Learning, Computer Vision, Machine Perception, Robotics, Feedback Control Systems, Design of Mechatronic Systems MANIPAL INSTITUTE OF TECHNOLOGY Manipal, India

Bachelor of Technology in Mechanical Engineering and Manufacturing Technology May 2017 WORK EXPERIENCE

Figueroa Lab @ GRASP LAB, University of Pennsylvania Philadelphia, PA Research Engineer June 2022 – Present

● Designed a pipeline for 6-DOF, occlusion-aware object tracking from segmented point cloud data for generating optimum grasp poses for novel objects using NVIDIA’s BundleSDF and Intel Realsense camera.

● Point cloud reconstruction using Log-GPIS and Open3D for robot perception and control.

● Created NeRF representation of indoor environment to encode obstacle geometry for vision-only robot navigation purposes.

● Designing a pipeline for thermal camera based human detection using deep recurrent neural network.

● Working on LiDAR-Depth Camera Sensor Fusion for SICK LiDAR Challenge 2023-2024. Rehabilitation Robotics Lab @ GRASP LAB, University of Pennsylvania Philadelphia, PA Research Intern March 2022 – June 2022

● Led a 10-member team in integrating off-the-shelf gaming wheels with force feedback for the budget friendly haptic TheraDrive system, aiding post-stroke rehabilitation.

● Utilized Arduino to calibrate bio-feedback sensors, enabling precise real-time patient data collection for the TheraDrive.

● Engineered a real-time feedback algorithm for dynamic game difficulty adjustment based on sensor data and patient game scores, enhancing user engagement and outcomes. PROJECTS

PICK AND PLACE CHALLENGE USING FRANKA ROBOTIC ARM, University of Pennsylvania

• Developed a software algorithm for the Panda robotic arm in Python to pick, orient and stack both static and moving blocks autonomously using ROS.

• Implemented arm's Forward and Inverse Kinematics and path planning algorithms like RRT and A* in the 7D joint space.

MACHINE PERCEPTION AND VISION, University of Pennsylvania

• Applied projective transformations and homographies in Python and OpenCV to project logos onto video sequence.

• Augmented Reality (AR) with AprilTags: Developed an AR application by solving the PnP, P3P and Procrustes problem.

• Structure from Motion (SfM): Implemented SfM pipeline to recover 3D transformation between two views using SIFT feature matching, 8-point algorithm with RANSAC filtering, camera pose estimation and point triangulation.

• 3D Reconstruction using Stereo Vision: Performed stereo rectification, block matching with SAD, SSD and ZNCC kernels, and left-right consistency check to build disparity map and reconstruct point cloud. MACHINE LEARNING PROJECTS

• People Counter using YOLO-NAS and SORT Object Tracking.

• Push Up Counter using YOLOv7 Pose Estimation.

• License Plate Detection and Recognition using YOLOv7.

• Waste Detection and Segmentation using YOLOv7.

• Car Velocity Calculation and Traffic Counting using YOLOv7 and Object Tracking.

• PPE Detection using YOLOv7.

• Built a Speech Recognition and Summarization System using vosk, pydub and transformers packages.

• Stock Price Predictor using yfinance package.

• mRNA Degradation Prediction using deep neural networks. SKILLS

Programming: Python, C++, MATLAB.

OS: Windows, Linux, macOS

Robotics & Machine Learning Tools: Pytorch, TensorFlow, Keras, Scikit-learn, fast.ai, Numpy, Pandas, PyBullet, Gazebo, RViz, ROS, OpenCV, Open3D, Simulink, Point Cloud Library, SAM, YOLO, CUDA, Photogrammetry. Software: Docker, Git, Arduino, Catia, Solidworks, COMSOL, ANSYS Fluent. Fabrication: CNC Milling, Laser cutting, 3D printing, soldering, circuit design.