I am an incoming Ph.D. student in Computer Science at Stony Brook University, advised by Prof. Chenyu You. My research broadly lies in computer vision, multimodal learning, and generative models.
My research focuses on building efficient multimodal and generative models that can understand, reason about, and interact with the physical world.
Efficient Multimodal and Generative Models: Improving visual and multimodal models through efficient representation learning, token reduction, and test-time mechanisms.
World Models and Embodied AI: Learning representations that support reasoning, planning, and interaction in dynamic environments.
Dynamic and 3D Visual Understanding: Studying event-based vision, 3D perception, and robust scene understanding under fast-changing conditions.
@inproceedings{hong2026regtoken,author={{<u>Cheng{-}Yao Hong</u>} and Wang, Yifan and Lin, Yuewei and You, Chenyu},title={Test-Time Registers as Global Priors for Tokenized Image Generation},booktitle={European Conference on Computer Vision (ECCV)},address={Malmö},year={2026},}
NeurIPS
Promptable 3-D Object Localization with Latent Diffusion Models
Cheng-Yao Hong , Li-Heng Wang , and Tyng-Luh Liu
In Advances in Neural Information Processing Systems 38 (NeurIPS)
@inproceedings{hong2025prompt,author={{<u>Cheng{-}Yao Hong</u>} and Wang, Li{-}Heng and Liu, Tyng{-}Luh},title={Promptable 3-D Object Localization with Latent Diffusion Models},booktitle={Advances in Neural Information Processing Systems 38 (NeurIPS)},address={San Diego},year={2025},dimensions={true},}
AAAI
Multimodal Promptable Token Merging for Diffusion Models
Cheng-Yao Hong , and Tyng-Luh Liu
In Association for the Advancement of Artificial Intelligence (AAAI)
@inproceedings{hong2025multimodal,author={{<u>Cheng{-}Yao Hong</u>} and Liu, Tyng{-}Luh},title={Multimodal Promptable Token Merging for Diffusion Models},booktitle={Association for the Advancement of Artificial Intelligence (AAAI)},address={Philadelphia},year={2025},dimensions={true},}
CVPR
Contrastive Learning for DeepFake Classification and Localization via Multi-Label Ranking
Cheng-Yao Hong , Yen-Chi Hsu , and Tyng-Luh Liu
In IEEE/CVF Computer Vision and Pattern Recognition Conference (CVPR)
@inproceedings{hong2024contrastive,author={{<u>Cheng{-}Yao Hong</u>} and Hsu, Yen{-}Chi and Liu, Tyng{-}Luh},title={Contrastive Learning for DeepFake Classification and Localization via Multi-Label Ranking},booktitle={{IEEE}/{CVF} Computer Vision and Pattern Recognition Conference (CVPR)},address={Seattle},year={2024},dimensions={true},}
ICCV
Attention Discriminant Sampling for Point Clouds
Cheng-Yao Hong , Yu-Ying Chou , and Tyng-Luh Liu
In IEEE/CVF International Conference on Computer Vision (ICCV)
@inproceedings{hong2023attention,author={{<u>Cheng{-}Yao Hong</u>} and Chou, Yu{-}Ying and Liu, Tyng{-}Luh},title={Attention Discriminant Sampling for Point Clouds},booktitle={{IEEE}/{CVF} International Conference on Computer Vision (ICCV)},address={Paris},year={2023},dimensions={true},}
TIP
ABC-Norm Regularization for Fine-Grained and Long-Tailed Image Classification
Yen-Chi Hsu* , Cheng-Yao Hong* , Ming-Sui Lee , Davi Geiger , and Tyng-Luh Liu
@article{HsuHLGL23,author={Hsu*, Yen{-}Chi and <u>{Cheng{-}Yao Hong}</u>* and Lee, Ming{-}Sui and Geiger, Davi and Liu, Tyng{-}Luh},title={ABC-Norm Regularization for Fine-Grained and Long-Tailed Image Classification},journal={{IEEE} Transactions on Image Processing (TIP)},volume={32},pages={3885--3896},year={2023},dimensions={true},}
Academic Services
I have served as a reviewer for CVPR, ICCV, ECCV, NeurIPS, ICLR, ICML, AAAI, and ICASSP, as well as for journals including TPAMI, TCYB, TNNLS, TNN, and Pattern Recognition.
🎉 
