Timezone: »
We introduce MOVE, a novel method to segment objects without any form of supervision. MOVE exploits the fact that foreground objects can be shifted locally relative to their initial position and result in realistic (undistorted) new images. This property allows us to train a segmentation model on a dataset of images without annotation and to achieve state of the art (SotA) performance on several evaluation datasets for unsupervised salient object detection and segmentation. In unsupervised single object discovery, MOVE gives an average CorLoc improvement of 7.2% over the SotA, and in unsupervised class-agnostic object detection it gives a relative AP improvement of 53% on average. Our approach is built on top of self-supervised features (e.g. from DINO or MAE), an inpainting network (based on the Masked AutoEncoder) and adversarial training.
Author Information
Adam Bielski (University of Bern)
Paolo Favaro (University of Bern)
More from the Same Authors
-
2022 Poster: Semi-supervised Vision Transformers at Scale »
Zhaowei Cai · Avinash Ravichandran · Paolo Favaro · Manchen Wang · Davide Modolo · Rahul Bhotika · Zhuowen Tu · Stefano Soatto -
2019 Poster: Emergence of Object Segmentation in Perturbed Generative Models »
Adam Bielski · Paolo Favaro -
2019 Spotlight: Emergence of Object Segmentation in Perturbed Generative Models »
Adam Bielski · Paolo Favaro -
2017 Poster: Deep Mean-Shift Priors for Image Restoration »
Siavash Arjomand Bigdeli · Matthias Zwicker · Paolo Favaro · Meiguang Jin -
2017 Spotlight: Deep Mean-Shift Priors for Image Restoration »
Siavash Arjomand Bigdeli · Matthias Zwicker · Paolo Favaro · Meiguang Jin