About Me
I am a Machine Learning Engineer specializing in 3D scene reconstruction, object detection, and underwater image processing. I have both academic and industry experience, with a publication in TNNLS and a granted patent.
Skills
- Deep Learning (PyTorch, TensorRT)
- 3D Scene Reconstruction (NeRF, Gaussian Splatting)
- Object Detection Optimization
- Image Processing (Underwater / Low-light conditions)
Work Experience
Changan Automobile Software Technology Co., Ltd. (Aug 2023 – Present)
- Performed 3D reconstruction using NeRF and 3D Gaussian methods.
- Improved PSNR from 23.14 → 27.82 on in-house datasets.
Kuaishou Technology (Intern) (Jun 2023 – Aug 2023)
- Optimized politically sensitive content strategies (+6% interception rate).
- Enhanced underage smoking detection accuracy to 90%.
Samsung Research China – Beijing (Intern) (Jul 2021 – Sep 2021)
- Implemented VPS model with AutoAugment & Copy-Paste on Cityscapes dataset.
Research
-
Large-Scale Modular Neural Inverse Rendering for Autonomous Driving (2023–)
Applied 3D Gaussian methods for large-scale autonomous driving scene reconstruction. -
Image Enhancement Guided Object Detection in Visually Degraded Scenes
Published in IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2023.
DOI: 10.1109/TNNLS.2023.3274926
Also granted patent: Image Enhancement Guided Object Detection Method under Adverse Imaging Conditions (ZL202210503164.0). -
National Key R&D Program – Underwater Multi-Sensor Fusion (2020–2023)
Improved underwater image detection accuracy by multi-sensor fusion and style-transfer techniques.
Education
- M.Sc. in Electronic Information (Control Engineering), USTB, 2020–2023
- B.Eng. in Automation, Yanshan University, 2015–2019
Honors & Awards
- Third Prize, National College Student Mathematics Competition (2016)
- Second Prize, Siemens Cup Intelligent Manufacturing Challenge (2018)