Using computer vision and deep learning to improve warehouse safety, accuracy, and efficiency.

• Applied position-based visual servoing control schemes to Kinova Mico robotic arm• Fused object detection and object tracking to improve tracking robustness and enable dynamic assignment of objects of interest• Located an object and segmented its point cloud from raw RGBD sensor data• Estimated a pre-grasping pose for a dynamic object using point cloud alignment and particle filter• Designed software pipeline using ROS and wrote ROS nodes in C++

• Programmed UR3 robot to automate extrinsic and intrinsic calibrations of camera rigs deployed in warehouses• Developed and refined proprietary camera calibration pipeline to enhance usability and calibration accuracy• Optimized image data processing to locate camera rotational center at subpixel accuracy• Generated maps for calibrated cameras to improve localization and estimation• Integrated the calibration software into a ROS package and documented the usage of the ROS package

• Developed xCapture, a scalable and network-based camera control software system that can control a large number of machine vision cameras. The system provides image live views, data capturing, raw image processing, and data playback. • Researched and prototyped the virtual head-mounted camera(VHC), a facial performance tracking system utilizing two motorized pan/tilt mirrors and machine vision camera • Developed motor controller for VHC to reach high precision control ±0.009◦ • Reverse engineered auto-zoom and focus motors for camera lenses used on VHC • Published VHC research project in SIGGRAPH 2015 poster session • Managed and advised on LSX3.2 project producing a radius of a 4M lighting reproduction light stage dome consists of over 1000 LSX3.2 lighting units. • Altered LSX3.2 circuit boards per customer’s request using Altium Designer software • Negotiated purchases and sourced parts for 1200 circuit boards and other parts for LSX3.2 project • Planned hardware production processes and quality control procedure • Engineered and prototyped apparatus for research projects • Programmed software for Microchip PIC microcontrollers used in lab equipment and research projects • Advised and trained hardware interns • Maintained documentation for projects • Light Stage X 3.2 Project; Computational Camera Project; Specular Object Scanner; Virtual Head-Mounted Camera Proeject • Mechatronics; Microcontroller; C/C++ and Python programming • Device and program prototyping; PCB Design

• Joined R&D team and collaborated with professional engineers on hydrodynamic retarder project• Tested and controlled hydrodynamic retarder and collected data in the laboratory• Diagnosed oil leak problem of hydrodynamic retarder prototype• Presented prototype retarder to potential customers with project members