Yan Liu

Adapted a personally developed offline BEV perception pipeline for internship use, integrating pretrained BEVFusion inference with nuScenes camera, LiDAR, radar, and ego-pose data. Implemented sensor coordinate transformations, radar velocity association, Kalman filtering, Hungarian matching, and 3-second constant-velocity prediction for collision-risk visualization. Validated sensor alignment, tracking, and prediction logic with 25 automated tests on synthetic mini-scenes.

Computer Vision / AI Engineering Intern at myRide (2026-06 – Present)

• Adapted a personally developed offline BEV perception pipeline for internship use, integrating pretrained BEVFusion inference with nuScenes camera, LiDAR, radar, and ego-pose data.
• Implemented sensor coordinate transformations, radar velocity association, Kalman filtering, Hungarian matching, and 3-second constant-velocity prediction for collision-risk visualization.
• Validated sensor alignment, tracking, and prediction logic with 25 automated tests on synthetic mini-scenes.

Calibration Algorithm Engineer, Part-time at Hunan Nuoyi Technology Co., Ltd. (2024-11 – 2025-03)

• Contributed C++/OpenCV/Eigen calibration modules for camera–laser projector geometry, distortion modeling, pose estimation, least-squares optimization, and outlier filtering.
• Calibrated and checked 16 laser projectors, achieving approximately 3 mm mean projection error on the display plane using focal-length-based geometric conversion; built residual-based diagnostics.

Calibration Algorithm Engineer, Full-time at Human Horizons (HiPhi) (2022-07 – 2024-04)

• Improved and maintained production C++/OpenCV end-of-line camera calibration modules deployed across 3 vehicle models and 2 manufacturing sites, covering target detection, PnP initialization, and nonlinear refinement.
• Converted calibration processing from single-threaded to multithreaded execution, reducing seven-camera computation time by approximately 50%, from 14 s to 7 s.
• Implemented reprojection-error checks; diagnosed factory-lighting-related detection failures using captured images and runtime logs, and validated corrective changes.
• Contributed to lane-based online camera orientation estimation using geometric constraints, lane filtering, and multi-frame validation.

Software Test Engineer Intern at IBM (2022-03 – 2022-06)

• Executed functional, regression, and API tests with Postman; validated data using MySQL, reproduced defects, and worked with developers to verify fixes.

M.S. in Electrical and Computer Engineering – University of Massachusetts Amherst (2025-01 – 2026-12)

B.S. in Electronic Information Engineering – Changchun University (2018-09 – 2022-06)