Tao Hu

Ommer Lab.

prof_pic.jpg

Computer Vision & Learning Group

Akademiestr. 7,Munich

Ludwig Maximilian University of Munich

I am a Postdoctoral Research Fellow with Björn Ommer in Ommer-Lab ( Stable Diffusion Lab ), focused on exploring the scalability and generalization ablity of diffusion model in the nxtaim project. I finished my PhD at VISLab, University of Amsterdam, supervised by Cees Snoek and Pascal Mettes. Closely coorporated with David W Zhang, Basura Fernando.

I am recruiting for Bachelor, Master and PhD supervision in Munich and globally. If you're interested in collaborating, feel free to send an email.

Open to discussion and collaboration, feel free to send an email.

Focused on introducing inductive bias into neural network to achieve data-efficiency by few-shot learning, generative model, etc. Have a conviction that generative modelling will be the future of discriminative modelling.

I'm on the job market. Feel free to contact me.

Publication | GitHub | CV(updated in Nov.2023) |
LinkedIn
| Research Note | Chat with me |Wechat |

Tao Hu's wechat

news

Jul 01, 2024 Two papers(including ZigMa) got accepted by ECCV! Also ZigMa: DiT-style Mamba-based diffusion models has been accepted as oral in ICML Workshop on Long Context Foundation Models (LCFM) :sparkles:
Jun 03, 2024 Give a talk at Adobe Research and A-Star to introduce our work about ZigMa.
Apr 14, 2024 Co-hosting the ECCV 2024 AVGenL: Audio-Visual Generation and Learning workshop.
Mar 29, 2024 I am honored to be invited as the Area Chair of AI for Content Creation Workshop @ CVPR 2024 and Efficient Large Vision Models Workshop @ CVPR 2024.
Dec 11, 2023 An editing method on latent flow matching is accepted by AAAI 2024:sparkles:

selected publications

  1. Diffusion Models and Representation Learning: A Survey
    Michael Fuest , Pingchuan Ma , Ming Gui , and 3 more authors
    In Arxiv , 2024
    The interplay between diffusion models and representation learning
  2. ZigMa: A DiT-style Zigzag Mamba Diffusion Model
    Vincent Tao Hu , Stefan Andreas Baumann , Ming Gui , and 4 more authors
    In ECCV , 2024
    a DiT-style Mamba-based diffusion models
  3. Boosting Latent Diffusion with Flow Matching
    Johannes S. Fischer , Ming Gui , Pingchuan Ma , and 4 more authors
    In ECCV , 2024
    flow matching for super-resolution
  4. Continuous, Subject-Specific Attribute Control in T2I Models by Identifying Semantic Directions
    Stefan Andreas Baumann , Felix Krause , Michael Neumayr , and 3 more authors
    In Arxiv , 2024
    Prompt Editing in T2I models
  5. DepthFM: Fast Monocular Depth Estimation with Flow Matching
    Ming Gui , Johannes S. Fischer , Ulrich Prestel , and 6 more authors
    In Arxiv , 2024
    An exploration of flow matching for blazing fast and zero-shot depth estimation
  6. Guided Flow Vision Transformer from Self-Supervised Diffusion Features
    Vincent Tao Hu , Yunlu Chen , Mathilde Caron , and 3 more authors
    In Submission , 2024
  7. Motion Flow Matching for Human Motion Synthesis and Editing
    Vincent Tao Hu , Wenzhe Yin , Pingchuan Ma , and 7 more authors
    In Submission , 2024
  8. Training Class-Imbalanced Diffusion Model Via Overlap Optimization
    Divin Yan , Lu Qi , Vincent Tao Hu , and 2 more authors
    In arxiv , 2024
  9. ./fm.png
    Latent Space Editing in Transformer-based Flow Matching
    Vincent Tao Hu , David W Zhang , Pascal Mettes , and 3 more authors
    In AAAI 2024. Also appear in ICML 2023 Workshop, New Frontiers in Learning, Control, and Dynamical Systems , 2024
  10. ToddlerDiffusion: Flash Interpretable Controllable Diffusion Model
    Eslam Mohamed BAKR , Liangbing Zhao , Vincent Tao Hu , and 3 more authors
    In Submission , 2024
  11. ./scribbleseg.png
    Generative Data Augmentation Improves Scribble-supervised Semantic Segmentation
    Jacob Schnell , Jieke Wang , Lu Qi , and 2 more authors
    In SyntaGen CVPR workshop , 2024
    Explore diffusion model for data augmention in segmentation task.
  12. ./fm-s2s.png
    Flow Matching for Conditional Text Generation in a Few Sampling Steps
    Vincent Tao Hu , Di Wu , Yuki M. Asano , and 4 more authors
    In EACL , 2024
    Flow Matching for text generation
  13. ./sgdm-why.png
    Self-Guided Diffusion Models
    Tao Hu* , David W Zhang* , Yuki M. Asano , and 2 more authors
    In CVPR , 2023
    A bridge between the community of self-supervised learning and diffusion models. Short version to appear in NeurIPS 2022 Workshop on Score-Based Methods and NeurIPS 2022 Workshop Self-Supervised Learning Theory and Practice.