Rui Huang

Rui Huang

Undergraduate Student

About Me

Hello! 👏 I’m a third-year undergraduate student in UESTC’s “Everest Project” Computer Top-Talent Experimental Class (2023–2027), majoring in Computer Science. Now, I’m interested in LLM SFT/RL and reasoning, image/video generation, and unified multimodal models. Earlier, I explored AI for smart grids and remote-sensing image fusion. You can find my resume or 中文简历 here.

⭐ I am eager to discuss potential collaborations and am actively seeking research internship opportunities(industry/academia),including onsite roles.I ‘m also seeking for 2027 fall PHD position. Please feel free to contact me via email: [huang_rui@std.uestc.edu.cn],[paulafixamiworali@gmail.com] or WeChat: huangrui_dby if you are interested.I warmly welcome your message and look forward to connecting!

🔥 News

  • 2024.10:  🏆 I am happy to be awarded the National Scholarship of 2025.

  • 2025.09:  🎉 We pre-released UniVA, first to unify Video Understanding/Segmentation/Editing/Generation into traceable multi-step workflows via Plan–Act agents, multi-level memory, and modular tools, plus UniVA-Bench!

  • 2025.07:  🎉 I’m honored to be invited to be a reviewer of AAAI 2026!

  • 2025.06:  🏆 I’m so happy to be 1/30 of SenseTime Scholarship 2025! This is the greatest praise I have received so far.

  • 2025.06:  🎉 We pre-released Diffusion Dataset Condensation on arXiv, accelerating the pre-training of diffusion models by 100x!!!. The entire project used hundreds of H100/A100!!

  • 2025.06:  🎉 One paper has been accepted by IEEE Transactions on Industrial Informatics, completed during my first research internship in AI for smart grid.

  • 2025.03:  🎉 We released CoT Image, The first to introduce CoT-style strategies into image generation, it has attracted widespread attention in the community and received over 790+ stars!!

  • 2024.12:  🎉 Our paper Wavelet-Assisted Multi-Frequency Attention Network for Pansharpening has been accepted by AAAI 2025 and selected for Oral Presentation. See you in Philadelphia!!!!

  • 2024.10:  🏆 I am happy to be awarded the National Scholarship of 2024 and the Gratitude Scholarship for Modern Scientists.

  • 2024.09:  🎉 One paper has been accepted by IEEE Transactions on Sustainable Energy, completed during my first research internship in AI for smart grid.

  • 2024.08:  🎉 I am delighted to be going to Cambridge as a visiting student.

  • 2023.12:  🌱 The starting point of my academic journey.

📝 Selected Publications ＆ Preprints

Under Review
D2C

🔥Diffusion Dataset Condensation: Training Your Diffusion Model Faster with Less Data

Rui Huang* , Shitong Shao* , Zikai Zhou, Pukun Zhao, Hangyu Guo, Tian Ye, Lichen Bai, Shuo Yang, Zeke Xie†

[PDF] [Github]

TL;DR: Proposed D²C: Diffusion Dataset Condensation for diffusion models, enabling 100× faster training with 0.8%–4% data via sample selection and semantic enhancement; trained on hundreds of A800/H100 GPUs.

AAAI 2025 Oral
D2C

🔥Wavelet-Assisted Multi-Frequency Attention Network for Pansharpening

Jie Huang*, Rui Huang*, Jinghao Xu, Siran Pen, Yule Duan, Liangjian Deng†

[PDF] [Github]

TL;DR: Proposed WFANet for image fusion, combining wavelet transformation with attention, achieving SOTA on multiple datasets.

Under Review
D2C

🔥Can We Generate Images with CoT? Let’s Verify and Reinforce Image Generation Step by Step

Ziyu Guo*, Renrui Zhang*†, Chengzhuo Tong*, Zhizheng Zhao*, Rui Huang, Haoquan Zhang, Manyuan Zhang, Jiaming Liu, Shanghang Zhang, Peng Gao, Hongsheng Li, Pheng-Ann Heng

[PDF] [Github🌟790+]

TL;DR: Proposed CoT-Image with step-wise reasoning and novel reward models (PARM/PARM++), improving autoregressive image generation by 24% via test-time verification and preference alignment.

Under Review
D2C

🔥UniVA: Universal Video Agents towards Next-Generation Video Intelligence

Zhengyang Liang*, Daoan Zhang*, Huichi Zhou, Rui Huang, Bobo Li, Shengqiong Wu, Yuechen Zhang, Xiaohan Wang, Jiebo Luo, Lizi Liao, Hao Fei

[PDF] [HomePage]

TL;DR: UniVA unifies understanding/segmentation/editing/generation into traceable multi-step video workflows via Plan–Act agents, multi-level memory, and modular tools, plus UniVA-Bench.

Under Review
D2C

Complementary Online Learning Network for Probabilistic Load Forecasting Against Extreme Weather

Rui Huang, Pengfei Zhao, Di Cao†, Weihao Hu, Qi Huang, Zhe Chen

[PDF] [Github]

TL;DR: Proposed the Complementary Online Learning Network (COLNet) with a Weather-aware gating mechanism for high precision probabilistic and point forecasting under extreme weather.

📖 Educations

  • 2023.09 - 2027.06, “Everest Project” Computer Top Talent Experimental Class, University of Electronic Science and Technology of China

🏆 Honors

  • [2025] SenseTime Scholarship（商汤奖学金） (award rate <0.1%; 30 undergrads China-wide)
  • [2025] National Scholarship（国家奖学金） (top recipient in the college)
  • [2024] National Scholarship（国家奖学金） (top recipient in the college)
  • [2024] Gratitude Scholarship for Modern Scientists （感恩中国近现代科学家助学金） (top 10 in school)
  • [2024] Excellence Scholarship — School of Computer Science, UESTC
  • [2024] First-Class Scholarship for Outstanding Students

🏅 Competition Awards

  • [2025] National Gold Award — National College Student Career Planning Competition（职规赛国金）🏅 (award rate <0.1%; first from UESTC to receive this award)
  • [2025] Provincial First Prize — China International College Students’ Innovation Competition （国创赛省一）🏅
  • [2025] Provincial Special Prize — “Challenge Cup” National College Students’ Extracurricular Academic and Technological Works Competition（挑战杯省特）🏅
  • [2024] National Third Prize — Five-Minute Research Presentation (5MRP) Competition🥉
  • [2023] Provincial First Prize — Huawei ICT Competition🏅（华为ICT省一） (rank 2rd in province)

💬 Talks ＆ Reports

🎓 Services

  • Reviewer for AAAI 2026
  • Co-Founder of the UESTC AI Club
  • Deputy captain for the “Embrace the Great, Pursue the Subtle” Social Practice Team（“致广大 尽精微”实践队）