Huijie Zhang

I am an Ph.D. student at University of Michigan, Ann Arbor, supervised by Prof. Qing Qu.

Previously, I obtained my Bachelor's degree in Mechanical Engineering from Huazhong university of Science and Technology, advised by Prof. Zhigang Wu; Master's degree in Mechanical Engineering and Electrical and Computer Engineering from University of Michigan, Ann Arbor, advised by Prof. Chad Jenkins

My research interests lie in generative model and diffusion model. Recently, my project is related the empirical and theoretical analysis of low-dimensional structures in diffusion model, and its applications related to training efficiency, controllable generation and privacy. My previous works are related to 3D vision, robotics manipulation and reinforcement learning.

[Updated in 10/2024]

Google Scholar  /  Github

[09/2024] Our work on LOCO Edit was accepted by NeurIPS 2024!

[05/2024] Our work on Diffusion Model Reproducibility was accepted by ICML 2024!

[02/2024] Our work on Multi-stage Diffusion Model was accepted by CVPR 2024!

[10/2023] Our work on Diffusion Model Reproducibility was accepted by NeurIPS2023 Workshop, received best paper award !

[08/2022] Our work on TransNet was accepted by ECCV 2022 Workshop!

[07/2022] Our work on Clearpose was accepted by ECCV 2022!

[06/2022] Our work on Progresslabeller was accepted by IROS 2022!

In this work, we propose Shallow Diffuse, utilizing the low-dimensional subsapce in diffusion model to disentangle the watermarking and image generation process. Shallow Diffuse is both emprirically and theoretically demonstrated its robustness and consistency, outperform previous watermarking techniques.

In this work, we provide theoretical insights into the connection between diffusion model and subspace clustering. The connection shed light into the transition of diffusion model from memorization to generalization and the mechanism it breaks the curse of dimensionality.

We improve the understanding of the semantic space in diffusion model and propose LOCO Edit, an editing method achieving precise and disentangled image editing without additional training. The proposed method is also supported by theoretical justification and has nice properties: homogeneity, transferability, composability, and linearity.

We investigate an intriguing and prevalent phenomenon of diffusion models: given the same starting noise input and a deterministic sampler, different diffusion models often yield remarkably similar outputs. And reveal its relationship with diffusion model generalizability.

In this study, we significantly enhance the training and sampling efficiency of diffusion models through a novel multi-stage framework. This method divides the time interval into several stages, using a specialized multi-decoder U-net architecture that combines time-specific models with a common encoder for all stages.

We proposed TransNet, a two-stage pipeline that learns to estimate category-level transparent object pose using localized depth completion and surface normal estimation.

We collected a large-scale transparent object dataset with RGB-D and annotated poses. And we benchmarked transparent object depth completion and poes estimation on this dataset.

ProgressLabeller is an efficient 6D pose annotation method. It is also the first open source tools compatible with transparent object. It was implemented as a blender Add-on, more user-friendly for using.

This Project useed Deep Q-learning from demonstration to teach agent cutting trees in Minecraft Environment

This website has been inspired by Jon Barron.