R&D Engineer, Master Degree, Open Source
Currently working at:
Studied at:
Reviewer: The Journal of Supercomputing, Pattern Analysis and Applications, Journal of Real-Time Image Processing
Email: chenyu1998424 [at] gmail [dot] com / chairc1998 [at] 163 [dot] com Location: Jinan, Shandong province, China
Github: chairc |
2023.06 - Present IDDM Distributed Defect Generation
Model Development Language: Python
Project Description: This system is an open-source distributed deep learning
image generation model (AIGC).
Technology Stack: PyTorch + Distributed
Project Content: The IDDM distributed defect generation model provides
non-repetitive industrial defect images based on DDPM and DDIM algorithms. The model uses DDP
distributed training technology to communicate across 6 RTX 6000 GPUs, enabling large-memory
multi-GPU training. It has garnered
and
on GitHub.
Project Link: https://github.com/chairc/Industrial-Defect-Diffusion-Model
2023.06 - 2024.06 Wood Defect Online Detection
Platform Development Language: Java, Python
Project Description: Major Science and Technology Innovation Project of
Shandong Province.
Technology Stack: SpringBoot + MySQL + MyBatis + VUE + PyTorch + TensorRT +
Linux + Tomcat
Project Content: Completed the front-end and back-end systems of the wood
defect online detection platform based on the SpringBoot framework and related technologies. The
deep learning detection system is implemented using the PyTorch framework and the YOLOX
algorithm. Communication between SpringBoot and Python is achieved using Socket ports. The
project includes the development of a dataset generation model based on the diffusion model, as
well as one-stage deep learning detection. The accuracy of wood detection is over 90%, and the
detection speed can reach 100 frames per second.It has garnered
and
on GitHub.
Project Link: https://github.com/chairc/NRSD-MN-relabel
2023.03 - 2022.06 GomROS Community
Platform Development Language: Java
Project Description: The system is a B/S architecture developer community for
users of the GomROS operating system, providing a community platform for communication,
learning, and collaboration.
Technology Stack: SpringBoot + MySQL + MyBatis + Shiro + VUE + Linux +
Tomcat
Project Content: Developed and maintained the GomROS community platform based
on the SpringBoot framework and related technologies. Designed multi-user and multi-permission
with the Shiro security framework to ensure that users can access relevant transactions without
overstepping their authority. Implemented an online chat room function using WebSocket. Worked
on the deployment of the Tomcat server on Linux. The platform meets basic communication,
learning, and collaboration requirements.
Project Link: http://124.70.86.111:8096/bbs/index/index
2022.12 - 2023.03 GomROS Cloud
Platform Development Language: Java
Project Description: The system is a B/S architecture cloud platform management
system, providing status monitoring, information collection, and analysis services for robot
device management.
Technology Stack: SpringBoot + MySQL + MyBatis + Shiro + VUE + Linux +
Tomcat
Project Content: Developed and maintained the data collection and analysis of
the GomROS cloud platform based on the SpringBoot framework and related technologies. Achieved
rapid loading of AGV maps through buffer pool technology. Completed custom automated database
and table creation with splicing SQL. Implemented online monitoring and data collection for
robot devices, with an average daily data volume of 1000 entries. Achieved message sending and
receiving functionality between the front and back ends using WebSocket. Worked on the
deployment of the Tomcat server on Linux.
2022.07 - 2022.11 Sterilization Robot Target
Detection Development Language: C++, Python
Project Description: Target detection module for sterilization robots, using
computer vision for precise disinfection of door handles on embedded devices.
Technology Stack: PyTorch + TensorRT
Project Content: Developed the
target detection module for sterilization robots using the YOLOv5 algorithm based on PyTorch
code. This included functions such as detection data collection, labeling, training, etc.
Deployed the engine model on embedded devices TX2 using C++ language and TensorRT. Achieved a
detection rate of up to 60 frames per second in real-time detection, with an accuracy exceeding
97%.
I'm currently working on computer vision and researching industrial defect detection.
If you are interested in my research, project in github and any questions, please send me an email!
Football, coding, fitness, travel and so on.