Dongting Hu

prof_pic.jpg

Hi! I am Dongting Hu, and I will be joining Snap Research Creative Vision as a Research Scientist. I obtained my Ph.D. in Science in the School of Mathematics and Statistics at the University of Melbourne, advised by Prof. Mingming Gong, co-advised by Dr. Liuhua Peng and Dr. Tingjin Chu. I was a member of the Machine Learning and Reasoning (MLR) Group.

I interned at Snap Research Creative Vision, working on efficient text-to-image generative models with Dr. Yanwu Xu, Dr. Anil Kag, and Dr. Jian Ren.

My research interests lie in Generative AI and its applications to various computer vision tasks, with a particular focus on:

  • Large-scale Multimodal Generative Models
  • Efficient Generative AI for Edge Devices
  • 3D Reconstruction and Generation

news

Feb 27, 2025 SnapGen is accepted by CVPR 2025 as a highlight. It has also been reported by Snap Newsroom and TechCrunch.
Oct 07, 2024 One paper accepted by NeurIPS 2024.
Aug 19, 2024 I joined Snap Inc. as a research intern working on efficient text-to-image models.
Jul 14, 2023 One paper accepted by ICCV 2023.
Jul 09, 2022 One paper accepted by ECCV 2022.

selected publications

  1. snapgen++.png
    SnapGen++: Unleashing Diffusion Transformers for Efficient High-Fidelity Image Generation on Edge Devices
    D. Hu, A. Gupta, M. Gabidolla, and 12 more authors
    arXiv preprint, 2026
    PDF Website
  2. snapgen.png
    SnapGen: Taming High-Resolution Text-to-Image Models for Mobile Devices with Efficient Architectures and Training
    D. Hu, J. Chen, X. Huang, and 16 more authors
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
    Highlight
    PDF Website
  3. innout.png
    In-N-Out: Lifting 2D Diffusion Prior for 3D Object Removal via Tuning-Free Latents Alignment
    D. Hu, H. Fu, J. Guo, and 5 more authors
    In Advances in Neural Information Processing Systems (NeurIPS), 2024
    PDF Website
  4. mipvog.png
    Multiscale Representation for Real-Time Anti-Aliasing Neural Rendering
    D. Hu, Z. Zhang, T. Hou, and 3 more authors
    In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2023
    PDF Code
  5. conor.png
    Uncertainty Quantification in Depth Estimation via Constrained Ordinal Regression
    D. Hu, L. Peng, T. Chu, and 4 more authors
    In Proceedings of the European Conference on Computer Vision (ECCV), 2022
    PDF Code