Presentation

<
DMHomo: Learning Homography with Diffusion Models
SessionConsistent Text-to-Image
DescriptionDMHomo leverages diffusion-models to generate a realistic dataset for homography learning. We train the generative model using pseudo labels. Additionally, we introduce an iterative process that improves both the homography estimator and the diffusion models successively, producing a qualified dataset and a state-of-the-art network.
Authors
Haipeng Li
University of Electronic Science and Technology of China
Hai Jiang
Sichuan University
Ao Luo
Megvii Technology Limited
Ping Tan
Hong Kong University of Science and Technology
Haoqiang Fan
Megvii Technology Limited
Bing Zeng
University of Electronic Science and Technology of China
Shuaicheng Liu
University of Electronic Science and Technology of China
Presenter
Haipeng Li
University of Electronic Science and Technology of China, Megvii Technology
Event Type
Technical Paper
TimeMonday, 29 July 20242:40pm - 2:50pm MDT
LocationMile High 1
ACM Digital Library Journal Papers' PDFs
Conference Papers' PDFs
Session Time & Location
Sunday, 28 July 20246:00pm - 8:45pm MDTBluebird Ballroom
Monday, 29 July 20242:00pm - 3:30pm MDTMile High 1
Interest Areas
Research & Education
Recordings
Livestreamed
Recorded
Keywords
AI
Machine Learning
Rendering
Registration Categories
Full Conference
Full Conference Supporter
Virtual Access
Exhibitor Full Conference
Monday
Next PresentationNext PresentationBlue Noise for Diffusion Models