IODA: An input/output deep architecture for image labeling

J Lerouge (1, 2) , Romain Hérault (1, 2, 3, 4) , Clément Chatelain (3, 1, 2, 4) , Fabrice Jardin (5, 6) , Romain Modzelewski (5)

1 LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
2 DocApp - LITIS - Equipe Apprentissage
3 INSA Rouen Normandie - Institut national des sciences appliquées Rouen Normandie
4 NU - Normandie Université
5 QuantIF-LITIS - Equipe Quantification en Imagerie Fonctionnelle
6 CLCC Henri Becquerel - Centre de Lutte Contre le Cancer Henri Becquerel Normandie Rouen

In this article, we propose a deep neural network (DNN) architecture called Input Output Deep Architecture (IODA) for solving the problem of image labeling. IODA directly links a whole image to a whole label map, assigning a label to each pixel using a single neural network forward step. Instead of designing a handcrafted a priori model on labels (such as an atlas in the medical domain), we propose to automatically learn the dependencies between labels. The originality of IODA is to transpose DNN input pre-training trick to the output space, in order to learn a high level representation of labels. It allows a fast image labeling inside a fully neural network framework, without the need of any preprocessing such as feature designing or output coding. In this article, IODA is applied on both a toy texture problem and a real-world medical image dataset, showing promising results. We provide an open source implementation of IODA 12 .

Mots clés: deep neural network; sarcopenia; medical imaging; semantic segmentation; machine learning; image labeling; Deep learning architectures

Domaines: Apprentissage [cs.LG]; Imagerie médicale; Réseau de neurones [cs.NE]

J Lerouge, Romain Hérault, Clément Chatelain, Fabrice Jardin, Romain Modzelewski. IODA: An input/output deep architecture for image labeling. Pattern Recognition, 2015, ⟨10.1016/j.patcog.2015.03.017⟩. ⟨hal-01338014⟩