Lingzhi Zhang

I am a Research Scientist at Adobe, where I am in a team working on inpainting-related technologies, including Remove Tool (Photoshop 24.5) and Generative Fill (Photoshop beta 24.6 & Firefly). I recently obtained my PhD in Computer and Information Science at the University of Pennsylvnia, advised by Prof.Jianbo Shi. During my PhD, I also closely collaborated with Eli Shechtman, Zhe Lin, Connelly Barnes, Sohrab Amirghodsi, and Yuqian Zhou at Adobe through mulitiple internships.

Selected Publications

	Perceptual Artifacts Localization for Image Synthesis Tasks Lingzhi Zhang, Zhengjie Xu, Connelly Barnes, Yuqian Zhou, Qing Liu, He Zhang, Zhe Lin, Eli Shechtman, Sohrab Amirghodsi, Jianbo Shi ICCV 2023 arxiv / code / dataset We generalize Perceptual Artifacts Localization to ten diverse image synthesis, and shows promising accuracy. We also show the effectiveness of automatic artifacts fixing and quality assessment as downstream applications.
	Perceptual Artifacts Localization for Inpainting Lingzhi Zhang, Yuqian Zhou, Connelly Barnes, Sohrab Amirghodsi, Zhe Lin, Eli Shechtman, Jianbo Shi ECCV 2022 (Oral) arxiv / code / dataset We formulate a new task of learning perceptual artifacts localization on the inpainted images. We propose a high-quality labeled dataset, successuflly train a model to localize the perceptual artifacts, and demonstrate downstream practical applications to no-reference single image quality assessment and iterative fill.
	Inpainting at Modern Camera Resolution by Guided PatchMatch with Auto-Curation Lingzhi Zhang, Connelly Barnes, Kevin Wampler, Sohrab Amirghodsi, Eli Shechtman, Zhe Lin, Jianbo Shi ECCV 2022 arxiv / code / dataset We propose a hybrid deep learning and patch-based approach for inpainting at modern camera resolution (4K+).
	Fine-Grained Egocentric Hand-Object Segmentation: Dataset, Model, and Applications Lingzhi Zhang, Shenghao Zhou, Simon Stent, Jianbo Shi (* equal contribution) ECCV 2022 project page / arxiv / code / dataset We propose a fine-grained egocentric hand-object segmentation dataset and model, and demonstrate its usage to multiple downstream vision applications, such as activity recognition, hand-object reconstruction, and seeing through the hands in videos.
	Learning Diverse Object Placement by Inpainting for Compositional Data Augmentation Lingzhi Zhang, Tarmily Wen, Jie Min, David Han, Jianbo Shi ECCV 2020 arxiv We study the problem of common sense placement of visual objects in an image.
	Nested Scale-Editing for Conditional Image Synthesis Lingzhi Zhang, Jiancong Wang, Yinshuang Xu, Jie Min, Tarmily Wen, James C. Gee, Jianbo Shi (* equal contribution) CVPR 2020 arxiv We proposed an image synthesis approach that provides stratified navigation in the latent code space.
	Deep Image Blending Lingzhi Zhang, Tarmily Wen, Jianbo Shi WACV 2020 arxiv / code We propose a deep optimization-based blending algorithm.
	Multimodal Image Outpainting with Regularized Normalized Diversification Lingzhi Zhang, Jiancong Wang, Jianbo Shi WACV 2020 arxiv / code We study the problem of generating a set of realistic and diverse backgrounds when given only a small foreground region, which we formulate as image outpainting task.
	Neural Embedding for Physical Manipulations Lingzhi Zhang, Andong Cao, Rui Li, Jianbo Shi (* equal contribution) Machine Learning for Physical Science Workshop, NeurIPS, 2019 arxiv Inspired by the properties of grid cells in mammalian brains, we build a generative model that enforces a normalized pairwise distance constraint between the latent space and output space to achieve data-efficient discovery of output spaces.