Guangze Zheng

Research assistant in Vision4Robotics

Tongji University

Biography

Guangze Zheng is pursuing the B.E. degree with the School of Mechanical Engineering (mainly engaged in Computer Science), Tongji University, Shanghai, China. As an undergraduate, He has won 12 scholarships. He is also a research assistant in Vision4Robotics Group, and his supervisor is Prof. Changhong Fu.

Download my resumé.

Interests

Deep learning
Robotics
Visual object tracking

Education

BEng in Mechanical Engineering (mainly engaged in Computer Science), 2018-2022

Tongji University

Experience

Research assistant

Vision4Robotics, Tongji University

Nov 2019 – Jun 2022 Shanghai

Supervisor: Prof. Changhong Fu

Develop real-time and robust aerial visual tracking methods
Explore unmanned aerial vehicle (UAV) tracking strategies in low-light environment
Introduce visual object tracking into unmanned aerial manipulator (UAM)

Research assistant

UAV Lab, Tsinghua University

Jul 2021 – Aug 2021 Beijing

Supervisor: Prof. Geng Lu

Perform visual tracking for UAM indoor flying
Design and record a UAM tracking benchmark

Core member

Zeal Eco-Power Racing Vehicle Team, Tongji University

Oct 2019 – Jun 2022 Shanghai

Supervisor: Prof. Diming Lou

Develop embedded equipment for racing vehicles
Design software for vehicle state visualization and driving assistance

Projects

Developing Real-Time and Robust Aerial Tracker on a Single CPU

Unmanned aerial vehicle (UAV) based visual tracking has been confronted with numerous challenges, e.g., object motion and occlusion. These challenges generally introduce unexpected mutations of target appearance and result in tracking failure.

Siamese Object Tracking for Unmanned Aerial Manipulator

This work introduces Siamese network into unmanned aerial manipulator (UAM) tracking, which is a model-free method and can be applied to track arbitrary objects. Considering severe object scale variation faced in practical UAM tracking application scenarios, a scale-aware Siamese network, i.

UAV Object Tracking in Low-Light Environment

Recent years have witnessed the fast evolution and promising performance of the convolutional neural network (CNN)-based trackers, which aim at imitating biological visual systems. However, current CNN-based trackers can hardly generalize well to low-light scenes that are commonly lacked in the existing training set.

Transformer Enhancer for UAV Tracking

Most previous progress in object tracking is realized in daytime scenes with favorable illumination. State-of- the-arts can hardly carry on their superiority at night so far, thereby considerably blocking the broadening of visual tracking-related unmanned aerial vehicle (UAV) applications.

Featured Publications

Guangze Zheng, Changhong Fu, Junjie Ye, Bowen Li, Geng Lu, Jia Pan

September 2021 Submitted to ICRA, 2022

Robust Siamese Object Tracking for Unmanned Aerial Manipulator

This work introduces Siamese network to unmanned aerial manipulator (UAM) tracking, and designs a scale attention strategy to deal with scale problems in UAM tracking applications.

Junjie Ye, Changhong Fu, Zi'ang Cao, Shan An, Guangze Zheng, Bowen Li

September 2021 Submitted RA-L, 2022

Tracker Meets Night: A Transformer Enhancer for UAV Tracking

This work proposes a transformer-based low-light image enhancer especially for UAV tracking.

Junjie Ye, Changhong Fu, Guangze Zheng, Zi'ang Cao, Bowen Li

March 2021 In IROS, 2021

DarkLighter: Light Up the Darkness for UAV Tracking

This work proposes a CNN-based low-light image enhancer for UAV tracking.

Guangze Zheng, Changhong Fu, Junjie Ye, Fuling Lin, Fangqiang Ding

November 2020 In ICRA, 2021

Mutation Sensitive Correlation Filter for Real-Time UAV Tracking with Adaptive Hybrid Label

This work proposes a mutation sensitive correlation filter to cope with the severe change of target appearance in UAV tracking.

Contact

mmlp@tongji.edu.cn
+86 18800250135
Cao’an Highway, Shanghai, 201804