I am a senior research scientist at Adobe Research. where my main focus at Adobe is to design the workflow for image composition including high-quality object segmentation, image harmonization, object/portrait relighting, object/portrait shadow generation and inverse rendering for image editing using diffusion model.  I was a computer vision engineer working at Apple Vision Pro.  I did my PhD advised by Vishal M. Patel.   

[CV] [Linkedin][Google Scholar][Github][Twitter]

Research Intern Hiring for Summer 2025 :  Please feel free to contact me if you are interested in working with me on a research project  in computer vision, deep learning, and their real-world applications especially  generative AI for multi-model image editing and image composition. 
Adobe University collaboration: feel free to email me, if you share similar interests in editing for  images, videos and texts!



News:
01/25: OS-refinement and Relighting-GS got accepted in ICLR'25. 

12/24: Text2Relight got accepted in AAAI'25. 

07/24: Generative Portrait Shadow Removal got accepted in Siggraph Asia'24. 

07/24: Two paper got accepted in ECCV'24. 

02/24: Three paper got accepted in CVPR'24. Relightful Harmonization, Relighting Nerf and Object Stitch V2. 

07/23: Photoswap: Personalized subject swapping in images is accepted in NeurIPS'23.

07/23: Perceptual Artifacts Localization for Image Synthesis Tasks is accepted in ICCV'23.

03/23: Rethinking portrait matting with privacy preserving is accepted in IJCV
(International Journal of Computer Vision)

02/23: Three paper got accepted in CVPR'23 on Relighting, Harmonization and Shadow/reflection synthesis. 

 01/23: Interactive Portrait Harmonization got accepted in ICLR 2023. Check out the project page.

11/22: Select Subject Cloud shipped in Photoshop (Link)

11/22: Semi-supervised Domain Alignment Learning for Single Image Dehazing got accepted in IEEE Transections on Cybernetics.

07/22: One paper got accepted in ECCV'2022.

05/22: Best Paper Award !!  "Image De-Raining Using a Conditional Generative Adversarial Network" receives the 2022 IEEE Transactions on Circuits and Systems for Video Technology Best Paper Award.

03/22: Two paper got accepted at CVPR'2022.        1. Lite Vision Transformer with Enhanced Self-Attention        2. Boosting Robustness of Image Matting with Context Assembling and Strong Data Augmentation 

07/21: SSH: A Self-Supervised Framework for Image Harmonization is accepted in ICCV'2021

03/21: MGMatting is accepted in CVPR'2021.  Our dataset and code can be found in the link.

       Email: he.zhang92@rutgers.edu          or                           hezhan@adobe.com                
                    (not hezhang@adobe.com)

Publications:

2023:


2022&Before:

Please refer to my Google Scholar

Product Highlights:

At Adobe, He  is the main contributor for the following shipped features:


And also more features in other products. 

More news:

10/21: Harmonization Filter  is shipped in 2021 Adobe MAX

03/21: One paper is accepted in IEEE Transactions on Cybernetics. 

10/20:  Refine Hair (combined with Object-aware Refine Edge) is shipped at 2020 Adobe MAX where an automatic trimap generation procedure is proposed. PiXimperfect **Check out others amazing features shipped by our group in this  video PiXimperfect like amazing Sky replacement, amazing content-aware tracing.

10/20: One paper is accepted in WACV'2021.

06/20:   My first feature in Photoshop. Select Subject Portrait is shipped in 2020 Photoshop June Release.  Design an efficient framework for automatic  high-quality portrait segmentation (with matted hair details )  running on-device. [PiXimperfect] [PHLEARN]

11/19:  One paper 'Generative Adversarial Networks with Fusion-discriminator for Single Image Dehazing' is  accepted in AAAI'2020.  

Working Experiences:


Selected Academic Services:

Former Interns:

I have been fortunate to collaborate with  these great students: