Hi there! I’m an ECE Ph.D. student at Cornell University, and I’m honored to be advised by Mert Sabuncu and Qingyu Zhao. In Summer 2026, I was a research intern on the Adobe Firefly foundation model team, working on large-scale pre-training and post-training of generative models. Previously, I worked on computer vision at CMU and on AI for healthcare at Georgia Tech and Emory.
Research Interests: Generative modeling (diffusion / flow matching, image & video generation), multimodal reasoning, and foundation-model pre-/post-training (on-policy distillation, RL), with applications to medical imaging.
Publications
Under Review
When to Unpair: Regulating Pairing Dependence in Medical Visual In-Context Learning
Cheng Wan, Chenjun Li, and Qingyu Zhao
Under review, 2026
Under Review
Self-Contained Representation Alignment for Pixel-Space Diffusion Transformers
Cheng Wan and Qingyu Zhao
Under review, 2026
NeurIPS
PRISM-Bench: A Benchmark of Puzzle-Based Visual Tasks with CoT Error Detection
Cheng Wan, Yusu Qian, and Qingyu Zhao
Advances in Neural Information Processing Systems (NeurIPS), 2026
@inproceedings{wan2026prism,title={PRISM-Bench: A Benchmark of Puzzle-Based Visual Tasks with CoT Error Detection},author={Wan, Cheng and Qian, Yusu and Zhao, Qingyu},booktitle={Advances in Neural Information Processing Systems (NeurIPS)},year={2026},}
TMI
Anatomically Guided Latent Diffusion for Brain MRI Progression Modeling
@article{wan2026agldm,title={Anatomically Guided Latent Diffusion for Brain MRI Progression Modeling},author={Wan, Cheng and Jafrasteh, Bahram and Adeli, Ehsan and Zhang, Miaomiao and Zhao, Qingyu},journal={IEEE Transactions on Medical Imaging (TMI)},year={2026},note={Under Revision},}
CVPR
Swift Parameter-free Attention Network for Efficient Super-Resolution
@inproceedings{wan2024swift,title={Swift Parameter-free Attention Network for Efficient Super-Resolution},author={Wan, Cheng and Yu, Hongyuan and Li, Zhiqi and Chen, Yihang and Zou, Yajun and Liu, Yuqing and Yin, Xuanwu and Zuo, Kunlong},booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},year={2024},note={Oral & Winner Award @ NTIRE Workshop},}
Under Review
LongCoref-AV: Benchmarking Audio-Visual Identity Coreference in Long-Form Video
Zhexiao Xiong, Huijuan Huang, Philip Anastassiou, Cheng Wan, Jing Shi, Manuel Brack, Yuheng Li, Ajinkya Kale, and Nathan Jacobs
Under review, 2026
MIDL
Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning
Chenjun Li, Cheng Wan, Laurin Lux, Alexander Berger, Richard B. Rosen, Martin J. Menten, and Johannes C. Paetzold
@inproceedings{li2026svr,title={Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning},author={Li, Chenjun and Wan, Cheng and Lux, Laurin and Berger, Alexander and Rosen, Richard B. and Menten, Martin J. and Paetzold, Johannes C.},booktitle={Medical Imaging with Deep Learning (MIDL)},year={2026},note={Spotlight},}
MICCAI
WASABI: A Metric for Evaluating Morphometric Plausibility of Synthetic Brain MRIs
@article{jafrasteh2025wasabi,title={WASABI: A Metric for Evaluating Morphometric Plausibility of Synthetic Brain MRIs},author={Jafrasteh, Bahram and Peng, Wei and Wan, Cheng and Luo, Yimin and Adeli, Ehsan and Zhao, Qingyu},journal={Medical Image Computing and Computer Assisted Intervention (MICCAI)},year={2025},}
2026
Under Review
When to Unpair: Regulating Pairing Dependence in Medical Visual In-Context Learning
Cheng Wan, Chenjun Li, and Qingyu Zhao
Under review, 2026
Under Review
Self-Contained Representation Alignment for Pixel-Space Diffusion Transformers
Cheng Wan and Qingyu Zhao
Under review, 2026
NeurIPS
PRISM-Bench: A Benchmark of Puzzle-Based Visual Tasks with CoT Error Detection
Cheng Wan, Yusu Qian, and Qingyu Zhao
Advances in Neural Information Processing Systems (NeurIPS), 2026
@inproceedings{wan2026prism,title={PRISM-Bench: A Benchmark of Puzzle-Based Visual Tasks with CoT Error Detection},author={Wan, Cheng and Qian, Yusu and Zhao, Qingyu},booktitle={Advances in Neural Information Processing Systems (NeurIPS)},year={2026},}
TMI
Anatomically Guided Latent Diffusion for Brain MRI Progression Modeling
@article{wan2026agldm,title={Anatomically Guided Latent Diffusion for Brain MRI Progression Modeling},author={Wan, Cheng and Jafrasteh, Bahram and Adeli, Ehsan and Zhang, Miaomiao and Zhao, Qingyu},journal={IEEE Transactions on Medical Imaging (TMI)},year={2026},note={Under Revision},}
Under Review
LongCoref-AV: Benchmarking Audio-Visual Identity Coreference in Long-Form Video
Zhexiao Xiong, Huijuan Huang, Philip Anastassiou, Cheng Wan, Jing Shi, Manuel Brack, Yuheng Li, Ajinkya Kale, and Nathan Jacobs
Under review, 2026
MIDL
Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning
Chenjun Li, Cheng Wan, Laurin Lux, Alexander Berger, Richard B. Rosen, Martin J. Menten, and Johannes C. Paetzold
@inproceedings{li2026svr,title={Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning},author={Li, Chenjun and Wan, Cheng and Lux, Laurin and Berger, Alexander and Rosen, Richard B. and Menten, Martin J. and Paetzold, Johannes C.},booktitle={Medical Imaging with Deep Learning (MIDL)},year={2026},note={Spotlight},}
arXiv
Unifying Active Learning and Semi-Supervised Learning for Medical Image Segmentation
Bahram Jafrasteh, Cheng Wan, Heejong Kim, Johannes C. Paetzold, and Qingyu Zhao
@article{jafrasteh2026unifying,title={Unifying Active Learning and Semi-Supervised Learning for Medical Image Segmentation},author={Jafrasteh, Bahram and Wan, Cheng and Kim, Heejong and Paetzold, Johannes C. and Zhao, Qingyu},journal={arXiv preprint arXiv:2607.25014},year={2026},}
arXiv
Your Simulation Runs but Solves the Wrong Physics: PDE-Grounded Intent Verification for LLM-Generated Multiphysics Simulation Code
@article{song2026pde,title={Your Simulation Runs but Solves the Wrong Physics: PDE-Grounded Intent Verification for LLM-Generated Multiphysics Simulation Code},author={Song, Zhenghan and Liu, Yulong and Wan, Cheng and Li, Chenjun and Liu, Lingfu and Li, Yunyi and Yuan, Congcong},journal={arXiv preprint arXiv:2605.09360},year={2026},}
arXiv
MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text
Chenjun Li, Cheng Wan, Haomiao Chen, and Johannes C. Paetzold
@article{li2026meld,title={MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text},author={Li, Chenjun and Wan, Cheng and Chen, Haomiao and Paetzold, Johannes C.},journal={arXiv preprint arXiv:2605.06903},year={2026},}
2025
MICCAI
WASABI: A Metric for Evaluating Morphometric Plausibility of Synthetic Brain MRIs
@article{jafrasteh2025wasabi,title={WASABI: A Metric for Evaluating Morphometric Plausibility of Synthetic Brain MRIs},author={Jafrasteh, Bahram and Peng, Wei and Wan, Cheng and Luo, Yimin and Adeli, Ehsan and Zhao, Qingyu},journal={Medical Image Computing and Computer Assisted Intervention (MICCAI)},year={2025},}
ICASSP
A Multi-scenario Attention-based Generative Model for Personalized Blood Pressure Time Series Forecasting
Cheng Wan, Chenjie Xie, Longfei Liu, Dan Wu, and Ye Li
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2025
@inproceedings{wan2025multi,title={A Multi-scenario Attention-based Generative Model for Personalized Blood Pressure Time Series Forecasting},author={Wan, Cheng and Xie, Chenjie and Liu, Longfei and Wu, Dan and Li, Ye},booktitle={IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},year={2025},}
Neurocomputing
HyCTAS: Multi-Objective Hybrid Convolution-Transformer Architecture Search for Real-Time Image Segmentation
Hongyuan Yu, Cheng Wan, Xiyang Dai, Mengchen Liu, Dongdong Chen, Bin Xiao, Yan Huang, Yuan Lu, and Liang Wang
@article{yu2025hyctas,title={HyCTAS: Multi-Objective Hybrid Convolution-Transformer Architecture Search for Real-Time Image Segmentation},author={Yu, Hongyuan and Wan, Cheng and Dai, Xiyang and Liu, Mengchen and Chen, Dongdong and Xiao, Bin and Huang, Yan and Lu, Yuan and Wang, Liang},journal={Neurocomputing},year={2025},}
BCB
KindSleep: Knowledge-Informed Diagnosis of Obstructive Sleep Apnea from Oximetry
Micky C. Nnamdi, Wenqi Shi, Cheng Wan, J. Ben Tamo, Benjamin M. Smith, Chad A. Purnell, and May Dongmei Wang
ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (BCB), 2025
@inproceedings{nnamdi2025kindsleep,title={KindSleep: Knowledge-Informed Diagnosis of Obstructive Sleep Apnea from Oximetry},author={Nnamdi, Micky C. and Shi, Wenqi and Wan, Cheng and Tamo, J. Ben and Smith, Benjamin M. and Purnell, Chad A. and Wang, May Dongmei},booktitle={ACM Conference on Bioinformatics, Computational Biology, and Health Informatics (BCB)},year={2025},doi={10.1145/3765612.3767236},}
2024
CVPR
Swift Parameter-free Attention Network for Efficient Super-Resolution
@inproceedings{wan2024swift,title={Swift Parameter-free Attention Network for Efficient Super-Resolution},author={Wan, Cheng and Yu, Hongyuan and Li, Zhiqi and Chen, Yihang and Zou, Yajun and Liu, Yuqing and Yin, Xuanwu and Zuo, Kunlong},booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)},year={2024},note={Oral & Winner Award @ NTIRE Workshop},}
J-BHI
Advancing Sleep Disorder Diagnostics: A Transformer-based EEG Model for Sleep Stage Classification and OSA Prediction
Cheng Wan, Micky C. Nnamdi, Wenqi Shi, Benjamin Smith, Chad Purnell, and May Dongmei Wang
IEEE Journal of Biomedical and Health Informatics (J-BHI), 2024
@article{wan2024sleep,title={Advancing Sleep Disorder Diagnostics: A Transformer-based EEG Model for Sleep Stage Classification and OSA Prediction},author={Wan, Cheng and Nnamdi, Micky C. and Shi, Wenqi and Smith, Benjamin and Purnell, Chad and Wang, May Dongmei},journal={IEEE Journal of Biomedical and Health Informatics (J-BHI)},year={2024},note={BHI Oral, 8% acceptance rate},}
BIBM
ECG-LLM: Leveraging Large Language Models for Low-Quality ECG Signal Restoration
Longfei Liu, Guosheng Cui, Cheng Wan, Dan Wu, and Ye Li
IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2024
@inproceedings{liu2024ecg,title={ECG-LLM: Leveraging Large Language Models for Low-Quality ECG Signal Restoration},author={Liu, Longfei and Cui, Guosheng and Wan, Cheng and Wu, Dan and Li, Ye},booktitle={IEEE International Conference on Bioinformatics and Biomedicine (BIBM)},year={2024},}
arXiv
RITA: A Real-time Interactive Talking Avatars Framework
Wuxinlin Cheng*, Cheng Wan*, Yupeng Cao*, and Sihan Chen*
@article{cheng2024rita,title={RITA: A Real-time Interactive Talking Avatars Framework},author={Cheng, Wuxinlin and Wan, Cheng and Cao, Yupeng and Chen, Sihan},journal={arXiv preprint arXiv:2406.13093},year={2024},}
arXiv
Unveiling Linear Mode Connectivity of Re-basin from Neuron Distribution Perspective
@article{li2024nonconnectivity,title={Unveiling Linear Mode Connectivity of Re-basin from Neuron Distribution Perspective},author={Li, Zhiqi and Li, Zexi and Wu, Mian and Wan, Cheng and Wu, Chao},journal={arXiv preprint},year={2024},}
2023
EMBC
Effects of Pulse Transit Time and Pulse Arrival Time on Cuff-less Blood Pressure Estimation: A Comparison Study with Multiple Experimental Interventions
Chenjie Xie, Cheng Wan, Yishan Wang, Jinzhong Song, Dan Wu, and Ye Li
Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2023
@inproceedings{xie2023ptt,title={Effects of Pulse Transit Time and Pulse Arrival Time on Cuff-less Blood Pressure Estimation: A Comparison Study with Multiple Experimental Interventions},author={Xie, Chenjie and Wan, Cheng and Wang, Yishan and Song, Jinzhong and Wu, Dan and Li, Ye},booktitle={Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)},year={2023},doi={10.1109/EMBC40787.2023.10340548}}
Experience
Research InternAdobe (Firefly)May 2026 – Aug 2026
Large-scale pre-training of Firefly foundation models: data curricula and mixture pipelines over billions of image/video samples; profiled multi-node training bottlenecks to refine training recipes.
Post-training for image generation with on-policy distillation (OPD) and reinforcement learning.
Awards & Honors
CVPR NTIRE Efficient Super-Resolution Challenge — 1st place & Oral2024
