Weakly-Supervised Semantic Segmentation using Motion Cues

Abstract : Fully convolutional neural networks (FCNNs) trained on a large number of images with strong pixel-level annotations have become the new state of the art for the semantic segmentation task. While there have been recent attempts to learn FCNNs from image-level weak annotations , they need additional constraints, such as the size of an object , to obtain reasonable performance. To address this issue, we present motion-CNN (M-CNN), a novel FCNN framework which incorporates motion cues and is learned from video-level weak annotations. Our learning scheme to train the network uses motion segments as soft constraints, thereby handling noisy motion information. When trained on weakly-annotated videos, our method outperforms the state-of-the-art approach on the PASCAL VOC 2012 image segmentation benchmark. We also demonstrate that the performance of M-CNN learned with 150 weak video annotations is on par with state-of-the-art weakly-supervised methods trained with thousands of images. Finally, M-CNN substantially out-performs recent approaches in a related task of video co-localization on the YouTube-Objects dataset.
Type de document :
Communication dans un congrès
ECCV 2016 - European Conference on Computer Vision, Oct 2016, Amsterdam, Netherlands. Springer, 9908 (Part IV), pp.388-404, 2016, Lecture Notes in Computer Science. <10.1007/978-3-319-46493-0_24>
Liste complète des métadonnées


https://hal.archives-ouvertes.fr/hal-01292794
Contributeur : Thoth Team <>
Soumis le : mardi 2 août 2016 - 17:26:12
Dernière modification le : jeudi 4 mai 2017 - 16:42:48
Document(s) archivé(s) le : mardi 8 novembre 2016 - 20:45:09

Fichier

mcnn.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Pavel Tokmakov, Karteek Alahari, Cordelia Schmid. Weakly-Supervised Semantic Segmentation using Motion Cues. ECCV 2016 - European Conference on Computer Vision, Oct 2016, Amsterdam, Netherlands. Springer, 9908 (Part IV), pp.388-404, 2016, Lecture Notes in Computer Science. <10.1007/978-3-319-46493-0_24>. <hal-01292794v3>

Partager

Métriques

Consultations de
la notice

728

Téléchargements du document

538