I am Cheng Chen, a postdoctoral researcher at the University of Trento, Italy, working under the supervision of Prof. Nicu Sebe. I joined the university in September 2026.
I received my Ph.D. in Computer Science and Technology from the University of Electronic Science and Technology of China (UESTC) in June 2026, where I was advised by Prof. Jingkuan Song and Prof. Lianli Gao. I began my graduate studies at UESTC in 2020, after earning my bachelorβs degree in Safety Engineering from Southwest Petroleum University.
My research focuses on continual learning and multimodal AI, especially how vision-language models can acquire new capabilities while retaining what they have already learned. My recent work explores continual instruction tuning, parameter-efficient adaptation, and efficient modeling of long image-text contexts. More broadly, I am interested in building computer vision and multimodal systems that can adapt effectively as tasks and data evolve.
π₯ News
- 2026.09: Joined the University of Trento as a postdoctoral researcher.
- 2026.06: π Two papers were accepted by ECCV 2026.
- 2026.05: π One paper was accepted by ICML 2026.
- 2026.01: π Received SCF Annual Outstanding Student Paper Award.
- 2024.10: π One paper was accepted by NeurIPS 2024.
- 2023.10: π Received National Scholarship Award.
- 2023.07: π One paper was accepted by ACM Multimedia 2023.
- 2022.07: π One paper was accepted by IEEE Transactions on Image Processing (TIP).
- 2022.07: π One paper was accepted by ACM Multimedia 2022.
πΌ Academic Position
- 2026.09 - Present, Postdoctoral Researcher, University of Trento, Trento, Italy
π Education
- 2022.09 - 2026.06, Ph.D.,
University of Electronic Science and Technology of China, School of Computer Science and Engineering, Computer Science and Technology, Chengdu, China - 2020.09 - 2022.06, Master, University of Electronic Science and Technology of China, School of Computer Science and Engineering, Computer Science and Technology, Chengdu, China
- 2010.09 - 2014.06, Bachelor,
Southwest Petroleum University, School of Chemistry & Chemical Engineering, Security Engineering, Chengdu, China.
π Publications
[From One-to-One to Many-to-Many: Dynamic Cross-Layer Injection for Deep Vision-Language Fusion]
Cheng Chen, Yuyu Guo, Pengpeng Zeng, Jingkuan Song, Peng Di, Hang Yu, Lianli Gao
ECCV 2026
Keywords: Vision Fusion, Instruction Tuning, Multimodel Large Language Model
[VLZip: Unified Visual and Textual Compression for Interleaved Long-Context Modeling]
Yuqi Zhang, Cheng Chen(Equal First Author), Yuyu Guo, Wenjie Yang, Lingchen Meng, Peng Di, Hang Yu, Zuxuan Wu, Yu-Gang Jiang
ECCV 2026
Keywords: Ultra-long, interleaved image-text, Instruction Tuning, Multimodel Large Language Model
CoIN: A Benchmark of Continual Instruction tuNing for Multimodel Large Language Model
Cheng Chen, Junchen Zhu, Xu Luo, Hengtao Shen, Lianli Gao, Jingkuan Song.
NeurIPS 2024 Datasets and Benchmarks
Keywords: Continual Learning, Instruction Tuning, Multimodel Large Language Model
Towards Redundancy-Free Sub-networks in Continual Learning
Cheng Chen, Jingkuan Song, LianLi Gao, Heng Tao Shen
Arxiv 2023
Keywords: Continual Learning, Information Bottleneck, SVD
CUCL: Codebook for Unsupervised Continual Learning
Cheng Chen, Jingkuan Song, Xiaosu Zhu, Junchen Zhu, Lianli Gao, Hengtao Shen
ACM Multimedia 2023
Keywords: Continual Learning, Unsupervised Learning, Vector Quantization
Continual Referring Expression Comprehension via Dual Modular Memorization
Heng Tao Shen, Cheng Chen, Peng Wang, Lianli Gao, Meng Wang, Jingkuan Song
IEEE Transactions on Image Processing
Keywords: Continual Learning, Visual Grounding, Rehearsal
Class Gradient Projection For Continual Learning
Cheng Chen, Ji Zhang, Jingkuan Song, Lianli Gao
ACM Multimedia 2022
Keywords: Continual Learning, Gradient Porjection, SVD
πΌ Internships
- 2025.06 - 2026.01, Research Intern, Ant Group, Multimodal Large Language Models
Built a 280K+ token ultra-long-context dataset from videos and documents for long-range reasoning, proposed Cross-Layer Injection (CLI) for many-to-many vision-language fusion, and achieved state-of-the-art results on 18 benchmarks for fine-grained perception.
π₯ Honors and Awards
- 2025 Sichuan Computer Federation (SCF) Annual Outstanding Student Paper Award, SCF
- 2023 National Scholarship, UESTC
- 2023 Outstanding Graduate Student Award, UESTC
- 2023 Academic First Prize Scholarship, UESTC
- 2022 Academic Second Prize Scholarship, UESTC, UESTC
- 2021 Academic Second Prize Scholarship, UESTC, UESTC