Chang Xu
PhD Student · ECEO Lab, EPFL, Switzerland

I am a PhD student (2024–) at the ECEO Lab, EPFL, supervised by Prof. Devis Tuia (EPFL) and Prof. Sebastian Schemm (University of Cambridge). During the PhD, my research focuses on reconstructing high-resolution Earth system dynamics with physical consistency, to benefit applications such as the forecasting of extreme weather events.

Prior to EPFL, I received both my B.Sc. and M.Sc. degrees from the Electronic Information School, Wuhan University, under the supervision of Prof. Wen Yang, where I worked on object detection and localization with a focus on tiny object detection.

chang.xu@epfl.ch GitHub Google Scholar
Chang Xu
Recent News
2026
Aug 2026 ECCV Our work on Rule-Based Feature Interpretation for Extreme Earth Events is accepted as an oral presentation at the AICC Workshop at ECCV 2026.
Apr 2026 Award Our work exploring different ways of Topography-aware Weather Forecasting wins the Best Poster Presentation Award at Climate Informatics 2026.
Jan 2026 ICLR A paper on High-fidelity Climate Super-resolution (GeoFAR) is accepted at ICLR 2026. Looking forward to the trip to Brazil!
2025
Nov 2025 TPAMI Our systematic work on Oriented Object Detection Dataset, Benchmark, and Method is accepted at IEEE TPAMI.
Jun 2025 ICCV Our SMARTIES on Multi-Sensor Remote Sensing Foundation Model is accepted at ICCV 2025.
May 2025 TPAMI Our work aiming at Detecting Every Object with Events is accepted at IEEE TPAMI.
Apr 2025 TGRS Our work discussing how to Minimize Sample Redundancy for Label-efficient Aerial Object Detection is accepted at IEEE TGRS.
Feb 2025 ISPRS Our work on Unbiased Semi-supervised Small Object Detection is accepted at ISPRS JPRS.
2024
Jul 2024 TGRS Our paper pushing forward the Cross-view Geo-localization without Ground Truth is accepted at IEEE TGRS.
Jul 2024 ECCV Our ConGeo paper that learns Robust Geo-localization Across Ground View Variations is accepted at ECCV 2024.
Mar 2024 ISPRS Our work on Object Tracking under Distraction is accepted at ISPRS JPRS.
2023
Nov 2023 TGRS Our work on Aerial Object Tracking is accepted at IEEE TGRS.
Aug 2023 ISPRS Our work on Drone-based RGBT Tiny Person Detection is accepted at ISPRS JPRS.
May 2023 Award I got the Best Presentation Award at ICDIP 2023.
Apr 2023 EPFL I will be starting a summer internship at EPFL, Switzerland.
Mar 2023 CVPR Our work on Oriented Tiny Object Detection is accepted at CVPR 2023.
2022
Jul 2022 ECCV Our work on RFLA: Gaussian Receptive Field-based Label Assignment for Tiny Object Detection is accepted at ECCV 2022.
Jun 2022 ISPRS Our work on the Tiny Object Detection Dataset, Benchmark and Method is accepted at ISPRS JPRS.
2021
Apr 2021 CVPRW Our work on Tiny Object Detection in Aerial Images is accepted at CVPR 2021 Workshop.
Selected Publications

* Equal Contribution  ·  # Corresponding Author

Earth System Modeling
GeoFAR
GeoFAR: Geography-Informed Frequency-Aware Super-Resolution for Climate Data
Chang Xu#, Gencer Sumbul, Li Mi, Robin Zbinden, Devis Tuia
ICLR 2026
Paper Project
SMARTIES
SMARTIES: Spectrum-Aware Multi-Sensor Auto-Encoder for Remote Sensing Images
Gencer Sumbul, Chang Xu, Emanuele Dalsasso, Devis Tuia
ICCV 2025
Paper Project
Small Object Detection
AI-TOD-R
Oriented Tiny Object Detection: A Dataset, Benchmark, and Dynamic Unbiased Learning
Chang Xu, Ruixiang Zhang, Wen Yang#, Haoran Zhu, Fang Xu, Jian Ding, Gui-song Xia
IEEE TPAMI 2025
Paper Project
DCFL
Dynamic Coarse-to-Fine Learning for Oriented Tiny Object Detection
Chang Xu, Jian Ding, Jinwang Wang, Wen Yang#, Huai Yu, Lei Yu#, Gui-song Xia
CVPR 2023
Paper Code
RFLA
RFLA: Gaussian Receptive Field based Label Assignment for Tiny Object Detection
Chang Xu, Jinwang Wang, Wen Yang#, Huai Yu, Lei Yu, Gui-song Xia
ECCV 2022
Paper Code
NWD
Detecting Tiny Objects in Aerial Images: A Normalized Wasserstein Distance and A New Benchmark
Chang Xu*, Jinwang Wang*, Wen Yang#, Huai Yu, Lei Yu, Gui-song Xia
ISPRS JPRS 2022 · Top-5 cited since 2022
Paper Code
Minimizing Sample Redundancy
Minimizing Sample Redundancy for Label-efficient Object Detection in Aerial Images
Ruixiang Zhang*, Chang Xu*, Haoran Zhu, Fang Xu, Wen Yang#, Haijian Zhang, Gui-Song Xia
IEEE TGRS 2025
Paper Code
Multi-modal & Cross-modal Learning
DEOE
Detecting Every Object from Events
Haitian Zhang, Chang Xu, Xinya Wang, Bingde Liu, Guang Hua, Lei Yu, Wen Yang#
IEEE TPAMI 2025
Paper Code
RGBT
Drone-based RGBT Tiny Person Detection
Yan Zhang*, Chang Xu*, Wen Yang#, Guangjun He, Huai Yu, Lei Yu, Gui-song Xia
ISPRS JPRS 2023
Paper Project
ConGeo
ConGeo: Robust Cross-view Geo-localization across Ground View Variations
Li Mi*, Chang Xu*,#, Javiera Castillo-Navarro, Syrielle Montariol, Wen Yang, Antoine Bosselut, Devis Tuia
ECCV 2024
Paper Project
EM-CVGL
Learning Cross-view Visual Geo-localization without Ground Truth
Haoyuan Li, Chang Xu, Wen Yang#, Huai Yu, Gui-Song Xia
IEEE TGRS 2024
Paper Project
Education
2024 –
EPFL, Switzerland
PhD Student, Environmental Computational Science and Earth Observation (ECEO)
2023
EPFL, Switzerland
Exchange Master Student
2021 – 2024
Wuhan University, China
M.Sc. in Information and Communication Engineering
2020
University of Oxford, UK
Visiting Student
2017 – 2021
Wuhan University, China
B.Sc. in Electronic Information Engineering
Awards & Honors
2023
Best Presentation Award
ICDIP 2023
2021
Outstanding Graduate Thesis
Top 5% · Wuhan University
2020
Hubei Translation Final Contest
Second Prize · Hubei, China
2019
National Undergraduate Electronic Design Contest (Hubei)
First Prize · Hubei, China