Learning Video Object Segmentation with Visual Memory

Pavel Tokmakov 1 Karteek Alahari 1 Cordelia Schmid 1
1 Thoth - Apprentissage de modèles à partir de données massives
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann
Abstract : This paper addresses the task of segmenting moving objects in unconstrained videos. We introduce a novel two-stream neural network with an explicit memory module to achieve this. The two streams of the network encode spatial and temporal features in a video sequence respectively, while the memory module captures the evolution of objects over time. The module to build a "visual memory" in video, i.e., a joint representation of all the video frames, is realized with a convolutional recurrent unit learned from a small number of training video sequences. Given a video frame as input, our approach assigns each pixel an object or background label based on the learned spatio-temporal features as well as the "visual memory" specific to the video, acquired automatically without any manually-annotated frames. The visual memory is implemented with convolutional gated recurrent units, which allows to propagate spatial information over time. We evaluate our method extensively on two benchmarks, DAVIS and Freiburg-Berkeley motion segmentation datasets, and show state-of-the-art results. For example, our approach outperforms the top method on the DAVIS dataset by nearly 6%. We also provide an extensive ablative analysis to investigate the influence of each component in the proposed framework.
Type de document :
Communication dans un congrès
ICCV - IEEE International Conference on Computer Vision, Oct 2017, Venice, Italy. 2017
Liste complète des métadonnées

Littérature citée [45 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01511145
Contributeur : Thoth Team <>
Soumis le : jeudi 10 août 2017 - 10:38:48
Dernière modification le : vendredi 11 août 2017 - 11:33:48

Fichier

1148.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01511145, version 2
  • ARXIV : 1704.05737

Collections

Citation

Pavel Tokmakov, Karteek Alahari, Cordelia Schmid. Learning Video Object Segmentation with Visual Memory. ICCV - IEEE International Conference on Computer Vision, Oct 2017, Venice, Italy. 2017. 〈hal-01511145v2〉

Partager

Métriques

Consultations de
la notice

341

Téléchargements du document

256