Pengcheng Liu

Forward-thinking and insightful experts with diversified skills in software system and machine learning. Experienced leader with multiple successful projects from idealization to production with varies project size from few people to hundreds of people. Exceptional problem-solving abilities both in team-oriented and self-motivated settings. Highly articulate with strong communication and interpersonal skills combined with deep technical expertise and innovated thinking capabilities.

Artificial Intelligence, Model Context Protocol, Analytics, Big Data, Cloud, Serverless, Microservice, OOD, Real-time Systems

Research and Development,  Agile Scrum,  Project Management, Product Management, Leadership

• 15 years of experience of software development
• 12 years of delivering commercial enterprise applications on premise and in the cloud
• 11 years of experience with big data analytics and distributed systems
• 14 years of experience with machine learning research and innovations
• 10 years of hands-on experience of machine learning algorithm development and production deployment
• 7 years of experience product and project management
• 5 years of experience of IoT platforms and solutions

• Medical Software System Project: High-energy fracture preoperative software application design
• Goal: Design and develop the clinical embraced software system to assist surgeon for preoperative planning
• Approach: Utilize Java and Java3D to implement the software system
• Cooperative trauma surgeons and orthopedics researchers perform software testing with real clinical cases
• The application contains image container and related analysis tool, 3D canvas with collision detection, scene graph structure, and underlying state-of-art image/surface segmentation and classification algorithms
• Achievement: Successfully restore fracture original anatomy under the designed application for 10 clinical cases
• The application is at the final stage to reach commercialization
• As chief application software architecture designer, team leader of the system development and main coordinator with orthopedics researchers and surgeons, during the application development
• Medical 3D Puzzle Solving Project: Automatically virtual 3D reconstruction of comminuted fracture
• Goal: Develop a novel computer algorithm for automated virtual reconstruction of comminuted fractures
• Approach: Novel 3D Puzzle solving algorithm using state-of-art surface matching, searching and registration techniques
• Achievement: Successfully reconstruct 10 clinical comminuted fractures with average 3 minutes
• One collaborative publication in biomedical article, one computer vision journal article in reviewing status
• 3D Surface Mesh Segmentation Project : Graph efficient surface mesh segmentation
• Goal: Develop an automated surface segmentation algorithm using the graph representation of the 3D surface
• Approach: Construct the surface as a graph and apply efficient graph cut segmentation method on the graph
• Achievement: Successfully divide surface into meaningful sub region
• One conference article in proceeding
• Embedded Vision system Project: Stereoscopic 3D Reconstruction using Motorized Zoom Lenses within an Embedded System
• Goal: Design a novel embedded stereoscopic 3D reconstruction system capable of estimating 3D positions of surfaces, which meant to act as a 3D sensing payload for a terrestrial robot
• Approach: Construct the hardware embedded system with a DSP running µclinux and an FPGA interacting to the system devices which consists of a CMOS camera pair and a pair of servo motors rotate each camera
• Program the system software of novel computer vision and image analysis tools such as camera calibration, stereo dense matching, and stereo reconstruction
• Achievement: (1) Automatic computation of the focus and exposure settings of the lens and camera sensor, (2) calibration of the system for various zoom settings and (3) Stereo reconstruction of free form objects
• One conference article was published in this project
• Archaeology Surface Registration Project: 3D Surface registration interface
• Goal: Implement a semi-automated surface registration interface means to stitch overlapped 3D surface together form a complete structure model
• Approach: Java, Java3D Implementation of the interface, surface matching and registration techniques, image stitching and texture remap
• Achievement: Align and registered 8 archaeology sites with total 122 individual scan data

Publication: Liu, P. and Willis, A. and Sui, Y., Stereoscopic 3D Reconstruction using Motorized Zoom Lenses within an Embedded System, SPIE Electronic Imaging, January 18–22, San Jose, California, Vol. 7251, pp.  72510W-72510W-12, 2009.

Publication: Thomas T.P., Anderson D.D., Willis A.R., Liu P.,   Frank M.C., Marsh J.L., and Brown T.D. A computational/experimental platform for investigating three-dimensional puzzle-solving of comminuted articular fractures. Computer Methods in Biomechanics and Biomedical Engineering.14 (3), pp. 263-270, 2011.

Publication: Liu, P. and Hewitt, N. and Waseem, G and Willis  A.R A system for 3D Reconstruction of Comminuted Tibial Plafond Bone Fractures. Computerized Medical Imaging and Graphics, vol. 2102.11684, 2021.

• Mandarin
• English