Yifu Zhang
实习生 at 百度
Wuxi City, Jiangsu, China
Summary
🤩
Rockstar🎓
Top SchoolYifu Zhang is a computer vision researcher and ML engineer with seven years of experience, currently a master's student at Huazhong University of Science and Technology and an intern at Baidu. He focuses on human-centric detection, multi-object tracking, and pose estimation, contributing to influential open-source projects like FairMOT and ByteTrack where he improved training pipelines, model architectures (HRNetV1→V2) and tracking–reID integration. His internship track record includes research roles at ByteDance and Microsoft Research China, blending academic rigor with production-oriented engineering. Based in Wuxi, he combines hardware background from an internship at Texas Instruments with summer CS study at UC Berkeley, reflecting a practical, systems-aware approach to applied vision problems.
8 years of coding experience
Summer Session, Computer Science, Summer Session, Computer Science at 美国加州大学伯克利分校硕士, Electrical, Electronics and Communications Engineering, 硕士, Electrical, Electronics and Communications Engineering at 华中科技大学
Github Skills (10)
multi-object-tracking10
computer-vision10
pytorch10
machine-learning10
python10
modeling9
trainings9
deep-learning9
net8
vnet8
Programming languages (2)
Github contributions (5)
FoundationVision/ByteTrack
Aug 2021 - Dec 2022
[ECCV 2022] ByteTrack: Multi-Object Tracking by Associating Every Detection Box
Role in this project:
ML Engineer
ML EngineerContributions:212 commits, 13 PRs, 40 pushes in 1 year 3 months
Contributions summary:Yifu contributed significantly to the `ifzhang/bytetrack` repository, which focuses on multi-object tracking. Their commits include modifications to the tracking framework, specifically focusing on the integration of DeepSORT and potentially other tracking algorithms like FairMOT. They made changes to the evaluator to incorporate the use of a DeepSORT re-identification model. The user updated scripts for improved performance and added new tracking configurations.
multi-object-trackingreal-timedeploymentpytorch
ifzhang/FairMOT
Apr 2020 - Dec 2021
[IJCV-2021] FairMOT: On the Fairness of Detection and Re-Identification in Multi-Object Tracking
Role in this project:ML Engineer
ML EngineerContributions:99 commits, 17 PRs, 95 pushes in 1 year 8 months
Contributions summary:Yifu made several contributions related to model training and evaluation for multi-object tracking. These included adding test configurations for the MOT20 dataset and modifying the training scripts and options. Additionally, the user updated the model architecture by changing HRNetV1 to HRNetV2 and incorporated fixes for uncertainty loss and dataset visualization bugs, showcasing a focus on improving model performance and training efficiency.
multi-object-trackingre-identificationone-shot-trackerjoint-detection-and-trackingreal-time
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.