Benteng Sun bio photo

Email

LinkedIn

Github

Google Scholar

👋 Welcome to My Personal Website!

About Me

Hi, I’m Benteng Sun (SMARK, 孙奔腾).

I’m an undergraduate student in Data Science (Mathematics Category) at Harbin Institute of Technology, Shenzhen (graduating in 2026).   My research centers on machine learning, sparse representation, and computational imaging, advised by Prof. Yongyong Chen.

I have published at ICLR and participated in top-tier competitions.   Currently, I am exploring efficient and controllable generation paradigms, aiming to build scalable frameworks for images, videos, and 3D content.

Feel free to reach out at SMARK2019@outlook.com.

Selected Publications

Spectral Compressive Imaging via Unmixing-driven Subspace Diffusion Refinement

Haijin Zeng*, Benteng Sun*, Yongyong Chen, Jingyong Su, Yong Xu
ICLR 2025 Accepted (Spotlight) • Top 4.79%📄 Paper | 💻 Code | 🌐 Project

PSR-SCI Pipeline
Developed a novel subspace diffusion refinement framework for spectral compressive imaging (SCI), recovering high-frequency details from minimal measurements. Achieved 38.14 dB PSNR on KAIST dataset with 10× speedup over existing diffusion models. Keywords: spectral imaging, sparsity modelling, diffusion refinement

Research Experience & Interests

Research Interests

My research bridges theoretical modelling and practical algorithm design, currently focusing on:

  • Multimodal Understanding & Generation   (cross-modal synthesis, language-guided visual generation, modality fusion)

  • Generative Modeling and Controllable Generation     (structure-guided diffusion, efficient sampling, scalable 2D/3D/video frameworks)

  • Computational Imaging & Low-Level Vision     (spectral imaging, MRI acceleration, physics-driven restoration, compressed sensing)

I am particularly interested in advancing new generation paradigms that integrate multimodal understanding, sparse structures, and efficient computation across 2D, 3D, and video modalities.

Competitions & Projects

  • The 19th “Challenge Cup” National Competition (First Prize)     Achieved over 95% accuracy in radar-infrared fusion object detection under limited-sample, noise-intensive conditions.

  • NTIRE Challenge (CVPR 2025)     Developed a novel diffusion-based image restoration system, compressing model parameters from 1.77B to 2.2M while retaining high visual quality.

  • MetaMusic: Cross-Modal Music-Driven Visual Generation (Top 1/83 in Freshman Project)     Designed a cross-modal generation pipeline using CLIP embeddings for music-to-image translation. 💻 Project

Honors & Awards

  • First Prize (National), “Challenge Cup” (MoE, PRC, 2024)
  • Meritorious Winner (Top 6%), Mathematical Contest in Modeling (COMAP, USA, 2024)
  • First Prize, 15th Chinese Mathematics Competition (2023)
  • Champion (1/83), Freshman Annual Research Project (HITSZ, 2023)

Scholarships

  • National Scholarship (Top 0.2%), MoE, PRC (2024-2025)
  • High-Level Innovation Award (Top 0.1%), HIT (2024, Awarded 50,000 RMB)

Feel free to explore my GitHub for open-source projects, or reach out via email for potential collaborations!

Thanks for visiting! 🌟