Motion-Refined DINOSAUR for Unsupervised Multi-Object Discovery

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Gong, Xinrui, Hahn, Oliver, Reich, Christoph, Singh, Krishnakant, Schaub-Meyer, Simone, Cremers, Daniel, Roth, Stefan
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916930245361664
author Gong, Xinrui
Hahn, Oliver
Reich, Christoph
Singh, Krishnakant
Schaub-Meyer, Simone
Cremers, Daniel
Roth, Stefan
author_facet Gong, Xinrui
Hahn, Oliver
Reich, Christoph
Singh, Krishnakant
Schaub-Meyer, Simone
Cremers, Daniel
Roth, Stefan
contents Unsupervised multi-object discovery (MOD) aims to detect and localize distinct object instances in visual scenes without any form of human supervision. Recent approaches leverage object-centric learning (OCL) and motion cues from video to identify individual objects. However, these approaches use supervision to generate pseudo labels to train the OCL model. We address this limitation with MR-DINOSAUR -- Motion-Refined DINOSAUR -- a minimalistic unsupervised approach that extends the self-supervised pre-trained OCL model, DINOSAUR, to the task of unsupervised multi-object discovery. We generate high-quality unsupervised pseudo labels by retrieving video frames without camera motion for which we perform motion segmentation of unsupervised optical flow. We refine DINOSAUR's slot representations using these pseudo labels and train a slot deactivation module to assign slots to foreground and background. Despite its conceptual simplicity, MR-DINOSAUR achieves strong multi-object discovery results on the TRI-PD and KITTI datasets, outperforming the previous state of the art despite being fully unsupervised.
format Preprint
id arxiv_https___arxiv_org_abs_2509_02545
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Motion-Refined DINOSAUR for Unsupervised Multi-Object Discovery
Gong, Xinrui
Hahn, Oliver
Reich, Christoph
Singh, Krishnakant
Schaub-Meyer, Simone
Cremers, Daniel
Roth, Stefan
Computer Vision and Pattern Recognition
Unsupervised multi-object discovery (MOD) aims to detect and localize distinct object instances in visual scenes without any form of human supervision. Recent approaches leverage object-centric learning (OCL) and motion cues from video to identify individual objects. However, these approaches use supervision to generate pseudo labels to train the OCL model. We address this limitation with MR-DINOSAUR -- Motion-Refined DINOSAUR -- a minimalistic unsupervised approach that extends the self-supervised pre-trained OCL model, DINOSAUR, to the task of unsupervised multi-object discovery. We generate high-quality unsupervised pseudo labels by retrieving video frames without camera motion for which we perform motion segmentation of unsupervised optical flow. We refine DINOSAUR's slot representations using these pseudo labels and train a slot deactivation module to assign slots to foreground and background. Despite its conceptual simplicity, MR-DINOSAUR achieves strong multi-object discovery results on the TRI-PD and KITTI datasets, outperforming the previous state of the art despite being fully unsupervised.
title Motion-Refined DINOSAUR for Unsupervised Multi-Object Discovery
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.02545