Author Archives: calvinadm
Technical report on analysing domain shift factors between videos and images for object detection.
Software release for Joint Calibration of the Ensemble of Exemplar-SVMs
Technical report on behavior discovery and alignment of articulated object classes from unstructured video
Release of TigDog dataset
Release of Zurich Summer dataset
Release of What’s the point dataset
New state-of-the-art on SIFT flow for semantic segmentation
We updated our BMVC paper: Joint Calibration for Semantic Segmentation. Using VGG16, our new results on SIFT Flow now outperform the state-of-the-art by 6% in the fully supervised setting, and by 10% in the weakly supervised setting. See our updated paper.