Junzhe Chen

410-710-5444 | junzhec12@gmail.com | jchen445@jh.edu

About Me

I'm a computer science researcher pursuing my Master's at Johns Hopkins University, with a focus on computer vision and deep learning. My passion lies in developing innovative solutions for visual foundation models, object detection, and model compression. With a background in software engineering from Beijing Jiaotong University and internships at tech companies like OPPO and Sense Technology, I blend theoretical knowledge with practical experience. I'm driven by the challenge of bridging academic research with real-world applications, always seeking to push the boundaries of what's possible in artificial intelligence.

Education

Research Interests

Computer Vision, Visual Foundation Models, Object Detection, Segmentation, Diffusion Models, Model Compression

Publications

Chen, J., et al. (2024). Adaptive quantization with mixed-precision based on low-cost proxy. In ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE.

Research Experience

Large decision-making model based on ODT
Dec 2023 - Jan 2024
Constructing a multimodal generalized decision-making large model using Online Decision Transformer, Multi-gate Mixture-of-Experts, and RL techniques.
Detection of extremely small targets under rain and fog
Oct 2023 - Jan 2024
Investigating YOLO, DETR, Dehazing, and Diffusion Model for small target detection in adverse weather conditions.
Efficient Adaptive Quantization with Mixed-Precision based on Low-Cost Proxy
Aug 2023 - Oct 2023
Developed a method using Hessian matrix trace for layer sensitivity determination and Pareto frontier approach for automated bit-width selection in mixed-precision quantization.

Professional Experience

Algorithm Engineer, Huizhi Square Technology Co., Ltd
Jul 2023 - Oct 2023
Researched SAM-related literature and deployed SAM models using quantitative techniques.
Deep Learning Intern, OPPO Communication Co., Ltd.
Nov 2022 - Jan 2022
Researched model compression techniques and conducted model accuracy and speed testing.
Computer Vision Algorithm Intern, Sense Technology Co., Ltd.
Sep 2022 - Nov 2023
Developed computer vision algorithms for motion capture and distance estimation using deep learning techniques.

Honors & Awards

Technical Skills

Python, PyTorch, ONNX, TensorRT