Zhi Tian - Doctoral Student at University of Adelaide

Zhi Tian

Doctoral Student at University of Adelaide

Australia

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Zhi Tian is a PhD-level computer vision researcher and engineer based in Australia with 12 years of experience focused on object detection, instance/semantic segmentation and OCR. Currently a doctoral student at the University of Adelaide, he combines rigorous academic research—documented in his publications—with hands-on engineering contributions to prominent open-source toolkits like AdelaiDet and an FCOS implementation. His work on core components (FCOSHead, CondInst), backbone refinements, ONNX export and utilities like NaiveGroupNorm demonstrates an ability to bridge research ideas into robust, production-ready model code. Prior roles at Malong Technologies, Shenzhen Institute of Advanced Technology and CUHK reflect applied research experience in both industry and labs. He’s notable for improving model post-processing, loss implementations and multi-scale/deformable conv support—skills that make him effective at taking state-of-the-art detection methods from paper to reproducible code.

12 years of coding experience

1 year of employment as a software developer

Bachelor's degree, Computer Science, Bachelor's degree, Computer Science at Sichuan University

Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Adelaide

Github Skills (11)

object-detection10

computer-vision10

pytorch10

machine-learning10

fc10

python10

instance-segmentation9

exports8

data-export8

exporter8

onnx8

Programming languages (5)

JavaC++ShellJupyter NotebookPython

Github contributions (5)

aim-uofa/AdelaiDet

Jan 2020 - Jun 2022

AdelaiDet is an open source toolbox for multiple instance-level detection and recognition tasks.

Role in this project:

ML Engineer

Contributions:1 review, 182 commits, 64 PRs in 2 years 5 months

Contributions summary:Zhi primarily contributes to the AdelaiDet repository, a toolbox for instance-level detection and recognition tasks, adding and modifying code related to the FCOS (FCOSHead) and CondInst object detection algorithms. These changes involve integrating new features, refactoring existing code, and fixing bugs within the core model components, including modifications to the FCOS outputs and associated loss functions. The contributions also include the addition of utilities such as `NaiveGroupNorm`, modification to the backbone architecture (DLA and ResNetLPF), and export the models to ONNX.

blendmaskdenseclfcostext-recognitiontoolbox

tianzhi0549/FCOS

Apr 2019 - Oct 2020

FCOS: Fully Convolutional One-Stage Object Detection (ICCV'19)

Role in this project:

Back-end Developer

Contributions:132 commits, 4 PRs, 58 pushes in 1 year 6 months

Contributions summary:Zhi's primary contribution involved adding and modifying FCOS (Fully Convolutional One-Stage Object Detection) functionality to the repository. This included implementing post-processing logic and loss functions within the model. The user also updated configurations and added a demo script for the FCOS model, indicating a focus on both implementation and usability. Further contributions included code modifications to support features like deformable convolution and multi-scale testing.

pytorchiccvimagenetbackbonedeep-learning

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.

Request Free Trial