Zhonghao Fang
School of Information and Electronics, Beijing University of Chemical Technology
School of Information and Electronics, Beijing Institute of Technology
No.5 Zhongguancun Street, Weigong Village, Haidian District, Beijing, China

Unsplashed background img 1


👋 About Me

I am now a M. S. student in Radar Research Lab, School of Information and Electronics, Beijing Institute of Technology starting from Autumn 2021.

I received the B. E. degree from School of Information and Electronics, Beijing University of Chemical Technology, Beijing, China, in 2021.

My research interests include Computer Vision, Auto Driving and Knowledge Distillation, with a focus on Horizontal / Oriented Object Detection and 3D Perception.

📑 Publications
An Adaptive Sample Assignment Strategy Based on Feature Enhancement for Ship Detection in SAR Images
Hao Shi, Zhonghao Fang, Yupei Wang* and Liang Chen
Remote Sensing (SCI Q2, IF=5.349)
paper
✨ Projects
An Oriented Ship Detector based RetinaNet By PyTorch

This is an open-source Oriented Ship Detector Rotation-RetinaNet implementation on Optical and SAR ship datasets (SSDD and HRSIC).

[Oriented-Ship-Detector-By-PyTorch] (Please visit repo for more details.)

[解读]
A Rotation Vehicle Detector based EfficientDet-D0 By PyTorch

This is an open-source Oriented Vehicle Detector Rotation-EfficientDet-D0 implementation on Optical Remote Sensing dataset (DOTA).

[Oriented-Vehicle-Detector-By-PyTorch] (Please visit repo for more details.)
An Oriented Multi-Class Object Detection Competition Solution

This is an Oriented Multi-Class Object Detector implementation on Optical Remote Sensing dataset (FAIR1M-1.5 + FAIR1M-2.0) based on Jittor Framework.

[Oriented-Object-Competition-With-Swin_Transformer] (Please visit repo for more details.)
A Horizontal Ship Detector Based RetinaNet By PyTorch

This is an open-source Horizontal Ship Object Detector Horizontal RetinaNet implementation on SAR Remote Sensing dataset (SSDD).

[Horizontal-Ship-Detector-By-PyTorch] (Please visit repo for more details.)

[解读]
📝 3D Perception Blogs
万字长文梳理BEVDet-Occupancy Network — 基于栅格网络的自驾感知算法

BEVDet-Occupancy Network是在BEVDet算法模型的基础上改进实现的Occupancy Network栅格占据网络，本文基于代码给出了bevdet-occ-r50-4d-stereo配置文件的文章解读。

Blog
万字长文理解纯视觉感知算法 — BEVFormer

BEVFormer是中稿ECCV 2022 的一篇论文。该篇论文提出了一个采用纯视觉(Camera)做感知任务的算法模型。该算法通过提取环视相机采集到的图像特征，并将提取的环视特征通过模型学习的方式转换到BEV空间，从而实现 3D 目标检测和地图分割任务，并取得了 SOTA 的效果。

Blog
LSS (Lift, Splat, Shoot) 论文+源码万字长文解析

LSS是一篇发表在ECCV 2020上有关自动驾驶感知方向的论文，具体子任务为object segmentation and map segmentation。

Blog
BEV视角下3D目标检测范式—BEVDet论文+源码解读

BEVDet：High-Performance Multi-Camera 3D Object Detection in Bird-Eye-View，该篇论文提出了一种基于BEV视角下的3D目标检测范式。

Blog
🏆 Awards
🎓 Education

B.E.degree from School of Information and Electronics, Beijing University of Chemical Technology, Beijing, China
Sep. 2017 - July 2021

M.S. in School of Information and Electronics, Beijing Institute of Technology, Beijing, China
Sep. 2021 - Present