Wujun Shao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Wujun Shao has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Chenzhou Cui
National Astronomical Observatories, Chinese Academy of Sciences
2 shared publications - Shirui Wei
Chinese Academy of Sciences
2 shared publications - Changhua Li
National Astronomical Observatories, Chinese Academy of Sciences
1 shared publication Chenzhou Cui
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China
1 shared publicationDongwei Fan
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China
1 shared publication- Dongwei Fan
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, People’s Republic of China;
1 shared publication
Latest Publications
- Open Access
- 19 May 2026
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 284 (2) , 45
A Value-added Physical Properties Catalog for Low-redshift Galaxies from DESI Legacy Imaging Surveys DR10
Galaxy physical properties—such as star formation rate (SFR), stellar mass, and gas-phase metallicity—are essential for population studies and evolutionary analyses. Deriving these quantities for billions of galaxies in modern imaging surveys presents significant challenges due to limited spectroscopy and the computational costs associated with traditional spectral energy distribution fitting. As a result, many galaxies in large photometric surveys still lack homogeneous property estimates. This study introduces a multimodal deep learning model that integrates optical imaging with photometric catalog features to estimate SFR, stellar mass, and oxygen abundance in low-redshift galaxies. The model incorporates a ResNet-based convolutional neural network to extract spatial information from multiband images and a multilayer perceptron that processes catalog-level photometric features, leveraging complementary constraints from morphology, surface brightness, and broadband colors. Trained on reference measurements from the MPA-JHU DR8 catalog, the model is optimized for efficient large-scale estimation. When applied to the DESI Legacy Imaging Surveys (LS) DR10, the model generates a value-added catalog containing physical property estimates for approximately 547 million galaxies with redshifts z ≤ 0.5. Validation through comparisons with independent catalogs and exploration of key scaling relations demonstrates that while the derived properties are not intended for precision measurements of individual objects, they effectively capture the dominant astrophysical trends necessary for ensemble studies. This catalog represents the first homogeneous set of photometry-based SFR, stellar mass, and metallicity estimates for DESI LS DR10, providing a vital resource for statistical studies of galaxies in the local Universe and facilitating comparisons with current and future spectroscopic surveys.
- 19 May 2026
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 284 (2) , 45
Open AccessA Value-added Physical Properties Catalog for Low-redshift Galaxies from DESI Legacy Imaging Surveys DR10
Galaxy physical properties—such as star formation rate (SFR), stellar mass, and gas-phase metallicity—are essential for population studies and evolutionary analyses. Deriving these quantities for billions of galaxies in modern imaging surveys presents significant challenges due to limited spectroscopy and the computational costs associated with traditional spectral energy distribution fitting. As a result, many galaxies in large photometric surveys still lack homogeneous property estimates. This study introduces a multimodal deep learning model that integrates optical imaging with photometric catalog features to estimate SFR, stellar mass, and oxygen abundance in low-redshift galaxies. The model incorporates a ResNet-based convolutional neural network to extract spatial information from multiband images and a multilayer perceptron that processes catalog-level photometric features, leveraging complementary constraints from morphology, surface brightness, and broadband colors. Trained on reference measurements from the MPA-JHU DR8 catalog, the model is optimized for efficient large-scale estimation. When applied to the DESI Legacy Imaging Surveys (LS) DR10, the model generates a value-added catalog containing physical property estimates for approximately 547 million galaxies with redshifts z ≤ 0.5. Validation through comparisons with independent catalogs and exploration of key scaling relations demonstrates that while the derived properties are not intended for precision measurements of individual objects, they effectively capture the dominant astrophysical trends necessary for ensemble studies. This catalog represents the first homogeneous set of photometry-based SFR, stellar mass, and metallicity estimates for DESI LS DR10, providing a vital resource for statistical studies of galaxies in the local Universe and facilitating comparisons with current and future spectroscopic surveys.
- Open Access
- 27 December 2025
- Published by Elsevier in Information Fusion
- Vol. 130
Deep learning-based astronomical multimodal data fusion: A comprehensive review
- 27 December 2025
- Published by Elsevier in Information Fusion
- Vol. 130
Open AccessDeep learning-based astronomical multimodal data fusion: A comprehensive review
- Open Access
- 24 May 2024
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 24 (6) , 065012
Astronomical Knowledge Entity Extraction in Astrophysics Journal Articles via Large Language Models
Astronomical knowledge entities, such as celestial object identifiers, are crucial for literature retrieval and knowledge graph construction, and other research and applications in the field of astronomy. Traditional methods of extracting knowledge entities from texts face numerous challenging obstacles that are difficult to overcome. Consequently, there is a pressing need for improved methods to efficiently extract them. This study explores the potential of pre-trained Large Language Models (LLMs) to perform astronomical knowledge entity extraction (KEE) task from astrophysical journal articles using prompts. We propose a prompting strategy called Prompt-KEE, which includes five prompt elements, and design eight combination prompts based on them. We select four representative LLMs (Llama-2-70B, GPT-3.5, GPT-4, and Claude 2) and attempt to extract the most typical astronomical knowledge entities, celestial object identifiers and telescope names, from astronomical journal articles using these eight combination prompts. To accommodate their token limitations, we construct two data sets: the full texts and paragraph collections of 30 articles. Leveraging the eight prompts, we test on full texts with GPT-4 and Claude 2, on paragraph collections with all LLMs. The experimental results demonstrate that pre-trained LLMs show significant potential in performing KEE tasks, but their performance varies on the two data sets. Furthermore, we analyze some important factors that influence the performance of LLMs in entity extraction and provide insights for future KEE tasks in astrophysical articles using LLMs. Finally, compared to other methods of KEE, LLMs exhibit strong competitiveness in multiple aspects.
- 24 May 2024
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 24 (6) , 065012
Open AccessAstronomical Knowledge Entity Extraction in Astrophysics Journal Articles via Large Language Models
Astronomical knowledge entities, such as celestial object identifiers, are crucial for literature retrieval and knowledge graph construction, and other research and applications in the field of astronomy. Traditional methods of extracting knowledge entities from texts face numerous challenging obstacles that are difficult to overcome. Consequently, there is a pressing need for improved methods to efficiently extract them. This study explores the potential of pre-trained Large Language Models (LLMs) to perform astronomical knowledge entity extraction (KEE) task from astrophysical journal articles using prompts. We propose a prompting strategy called Prompt-KEE, which includes five prompt elements, and design eight combination prompts based on them. We select four representative LLMs (Llama-2-70B, GPT-3.5, GPT-4, and Claude 2) and attempt to extract the most typical astronomical knowledge entities, celestial object identifiers and telescope names, from astronomical journal articles using these eight combination prompts. To accommodate their token limitations, we construct two data sets: the full texts and paragraph collections of 30 articles. Leveraging the eight prompts, we test on full texts with GPT-4 and Claude 2, on paragraph collections with all LLMs. The experimental results demonstrate that pre-trained LLMs show significant potential in performing KEE tasks, but their performance varies on the two data sets. Furthermore, we analyze some important factors that influence the performance of LLMs in entity extraction and provide insights for future KEE tasks in astrophysical articles using LLMs. Finally, compared to other methods of KEE, LLMs exhibit strong competitiveness in multiple aspects.