Yuqian Zhou (周宇乾)

  • I am a research scientist at Adobe. I am also the core researcher and engineer of Remove Tool (Photoshop 24.5), Generative Fill (Photoshop Beta 24.6, Adobe Firefly). I received my Ph.D. from UIUC and was advised by Prof. Mark Hasegawa-Johnson. Before that, I was supervised by Prof. Thomas Huang(1936-2020). I was a research assistent at Image Formation and Processing Group IFP led by Prof. Humphrey Shi of the University of Illinois at Urbana-Champaign (UIUC). I received my Bachelor and Mphil. degree from the Hong Kong University of Science and Technology(HKUST), working at Neuromorphic Interactive System Lab (NISL) supervised by Prof. Bertram Shi.

  • My research interests lie in computer vision, deep learning, human-computer interaction and multidisciplinary vision research. Specifically, I am actively conducting low-level vision research like image denoising, enhancement and image inpainting. I am engaging in applying image/data restoration and recognition theory to multiple reseach areas including optics and display, art and history, neuroscience, psychology and affect, medicine and health, even financial trading. Please see my list of publications for some examples.

Academic News

  • May., 2024 Generative Remove released in Adobe Lightroom.
  • Apr., 2024 Generative Fill with new Firefly Model 3 is released in PS Beta during Adobe MAX London 2024.
  • Feb., 2024. One paper accepted to CVPR’24.
  • Oct., 2023. Generative Fill / Remove Tool related features have been released in Photoshop Web, Adobe Express, Adobe Stock, Stardust, Adobe Illustrator and many Adobe Sneaks in Adobe MAX 2023.
  • 30 Sep., 2023. Auto Photo Expansion at Adobe Stock is released.
  • 23 Sep., 2023. Generative Fill officially in Photoshop 25.0
  • 15 July., 2023. One paper is accepted to ICCV’23.
  • 23 May., 2023. Generative Fill in Photoshop Beta 24.6, and in Adobe Firefly.
  • 23 May., 2023. Remove Tool released in Photoshop 24.5.
  • 27 Feb., 2023. Two papers are accepted to CVPR’23.
  • 21 Jan., 2023. One paper is accepted to ICLR’23 as top-5%.
  • 16 Aug., 2022. Two papers of inpainting are accepted to WACV’23.
  • 3 July., 2022. One paper of inpainting is accepted to ECCV’22 (Oral).
  • 4 Apr.. 2022. One paper related to Human-Computer Interaction is accepted by DIS’22.
  • 1 Mar., 2022. One paper of image vectorization is accepted to CVPR’22 (Oral).
  • 8 Nov., 2021. I passed my Ph.D. Thesis Defense. Having obtained offers from Huawei, Adobe, Apple and Amazon etc., I will join Adobe ART (Photoshop) as a research scientist.
  • 31 Aug., 2021. Awarded the outstanding reviewer of ICCV’21.
  • 11 Jun., 2021. SGL Vessel Segmentation paper is accepted to MICCAI’21. arXiv and the code is released here.
  • 29 Mar., 2021. Our reference-based inpainting work at CVPR’21, TransFill, is available in arXiv and the partial-code and the data will be released soon. More details at our project page.
  • 5 Mar., 2021. Our Retinal Vesssel Segmentation work appears in arXiv and the code is released here. The performance is the SOTA.
  • 28 Feb., 2021. Three papers are accepted to CVPR’21.
  • 1 Dec., 2020. One paper accepted to AAAI’21.
  • 28 Aug., 2020. We organized the first Challenge of Image Restoration for Under-Display Camera at ECCV ‘20. Challenge Report is released. Microsoft Project Page. Media Post.
  • 23 Feb, 2020. One paper accepted to CVPR’20.
  • 10 Feb, 2020. We organized the workshop and Tiny Object Detection (TOD) challenge on Real-World Recognition from Low-Quality Inputs (RLQ) in ECCV 2020. The program website is here. YouTube Recording is here
  • 10 Nov, 2019. Three papers are accepted to AAAI’20.
  • 27 Oct, 2019. Successfully organized the workshop and challenge on Real-World Recognition from Low-Quality Images and Videos (RLQ) in ICCV’19. The past program website is here. Thanks all the co-organizers and speakers. [Photo]
  • 27 Oct, 2019. We won the 3rd prize of Dunhuang Image Restoration Challenge at e-Heritage Workshop at ICCV 2019. [Certificate][Photo]
  • 22 July, 2019. One paper accepted to ICCV 2019 (Oral).[Video]
  • 11 June, 2019. One Journal paper accepted in PLoS ONE.
  • 16 April, 2019. Our Team has won the 1st place of the NTIRE 2019 Raw Image Denoising Track at CVPR’19! See the Tech Report.
  • 24 Feb, 2019. One paper is accepted to CVPR’19 (Oral).
  • 3 Nov, 2018. One paper and a student abstract paper are accepted in AAAI 2019.
  • 3 Jun, 2017. Our HKUST Team ‘Pose-dependent facial expression recognition system’ ranked 1st in the challenge of FERA2017. The associated paper is accepted in FG’17.

Tech Transfer

  • Photoshop / Firefly Generative Fill
  • Photoshop Generative Background
  • Photoshop Generative Similar
  • Photoshop / Firefly Generative Expand
  • Photoshop / Firefly Generative Fill with Reference Image
  • Photoshop Remove Tool
  • Lightroom Generative Remove
  • Adobe Stock Auto Expand

Selected Awards

  • Mar. 2020, 2020 Thomas and Margaret Huang Award for Graduate Research (Beckman, UIUC)
  • Apr. 2020, 5th (5/265) place of NTIRE2020 sRGB Image Denoising Challenge at CVPR’20.
  • Oct. 2019, 3rd (3/23) place of Dunhuang Image Restoration Challenge at ICCV’19.
  • Jun. 2019, 1st (1/220) place of NTIRE2019 Raw Image Denoising Challenge at CVPR’19.
  • Jun. 2017, 1st place of FERA2017 Action Unit Intensity Regression Challenge at FG’17.
  • Nov. 2015, 2015 Academic Achievement Medal of HKUST

Industrial Experience

  • Summer 2020, Fall 2020, Spring 2021, Summer 2021 @ Adobe Photoshop Team and Adobe Research
  • Summer 2019 @ Applied Science Group and Microsoft Research
  • Summer 2018 @ Megvii US Research
  • Fall 2018, Jump Trading Research Assistant

Reviewer

  • Conference: CVPR, ICCV, ECCV, NeruIPS, AAAI, ICLR, ICML, ACCV, WACV etc.
  • Journal: TIP, IEEE Transactions on Cybernetics, IEEE Transactions on Computational Survey, IEEE Transactions on Affective Computing etc.