Presentation
MTPano: Multi-Task Panoramic Scene Understanding via Label-Free Integration of Dense Prediction Priors
SessionDynamic Scenes
DescriptionMTPano is a robust multi-task foundation model for panoramic scene understanding. Trained via a label-free pipeline using perspective dense priors, it adopts a dual-stream architecture that resolves feature conflicts between depth, segmentation, and surface normal estimations, achieving state-of-the-art performance on spherical images.
Event Type
Technical Paper
TimeTuesday, 21 July 202610:45am - 10:55am PDT
LocationRoom 408 B
Digital Library
PDF
Session Time & Location
Sunday, 19 July 20266:00pm - 8:45pm PDTHall K
Tuesday, 21 July 202610:45am - 12:15pm PDTRoom 408 B
Artificial Intelligence/Machine Learning
Full Conference Supporter
Full Conference
