Keywords
Computer Science - Computer Vision and Pattern RecognitionComputer Science - Learning
Computer Science - Neural and Evolutionary Computing
Full record
Show full item recordOnline Access
http://arxiv.org/abs/1412.6885Abstract
The Convolutional Neural Network (CNN) has achieved great success in image classification. The classification model can also be utilized at image or patch level for many other applications, such as object detection and segmentation. In this paper, we propose a whole-image CNN regression model, by removing the full connection layer and training the network with continuous feature maps. This is a generic regression framework that fits many applications. We demonstrate this method through two tasks: simultaneous face detection & segmentation, and scene saliency prediction. The result is comparable with other models in the respective fields, using only a small scale network. Since the regression model is trained on corresponding image / feature map pairs, there are no requirements on uniform input size as opposed to the classification model. Our framework avoids classifier design, a process that may introduce too much manual intervention in model development. Yet, it is highly correlated to the classification network and offers some in-deep review of CNN structures.Date
2014-12-22Type
textIdentifier
oai:arXiv.org:1412.6885http://arxiv.org/abs/1412.6885
Collections
Related items
Showing items related by title, author, creator and subject.
-
Actes du CARI 2014 (Colloque africain sur la recherche en informatique et mathématiques appliquées)Université Badji Mokhtar - Annaba [Annaba] ( UBMA ); SUpervision of large MOdular and distributed systems ( SUMO ) ; Inria Rennes – Bretagne Atlantique ; Institut National de Recherche en Informatique et en Automatique ( Inria ) -Institut National de Recherche en Informatique et en Automatique ( Inria ) -LANGAGE ET GÉNIE LOGICIEL ( IRISA-D4 ) ; Institut de Recherche en Informatique et Systèmes Aléatoires ( IRISA ) ; CentraleSupélec-Université de Rennes 1 ( UR1 ) ; Université de Rennes ( UNIV-RENNES ) -Université de Rennes ( UNIV-RENNES ) -Télécom Bretagne-Institut National des Sciences Appliquées ( INSA ) -Institut National de Recherche en Informatique et en Automatique ( Inria ) -École normale supérieure - Rennes ( ENS Rennes ) -Centre National de la Recherche Scientifique ( CNRS ) -Université de Bretagne Sud ( UBS ) -CentraleSupélec-Université de Rennes 1 ( UR1 ) ; Université de Rennes ( UNIV-RENNES ) -Université de Rennes ( UNIV-RENNES ) -Télécom Bretagne-Institut National des Sciences Appliquées ( INSA ) -Institut National de Recherche en Informatique et en Automatique ( Inria ) -École normale supérieure - Rennes ( ENS Rennes ) -Centre National de la Recherche Scientifique ( CNRS ) -Université de Bretagne Sud ( UBS ) -Institut de Recherche en Informatique et Systèmes Aléatoires ( IRISA ) ; CentraleSupélec-Université de Rennes 1 ( UR1 ) ; Université de Rennes ( UNIV-RENNES ) -Université de Rennes ( UNIV-RENNES ) -Télécom Bretagne-Institut National des Sciences Appliquées ( INSA ) -Institut National de Recherche en Informatique et en Automatique ( Inria ) -École normale supérieure - Rennes ( ENS Rennes ) -Centre National de la Recherche Scientifique ( CNRS ) -Université de Bretagne Sud ( UBS ) -CentraleSupélec-Université de Rennes 1 ( UR1 ) ; Université de Rennes ( UNIV-RENNES ) -Université de Rennes ( UNIV-RENNES ) -Télécom Bretagne-Institut National des Sciences Appliquées ( INSA ) -École normale supérieure - Rennes ( ENS Rennes ) -Centre National de la Recherche Scientifique ( CNRS ) -Université de Bretagne Sud ( UBS ); Laboratoire d'Analyse Numérique et Informatique [Sénégal] ( LANI ) ; Université Gaston Berger de Saint-Louis Sénégal ( UGB ); Mokhtar Sellami; Eric Badouel; Moussa Lo; Sellami , Mokhtar; Badouel , Eric; Lo , Moussa (HAL CCSDInria, 2014-10-17)International audience
-
An Intelligent Mediating Model for Collaborative e-Learning Management SystemsAkanbi Caleb Olufisoye; Adagunodo E Rotimi (IJCSI Press, 2011-07-01)E-learning management systems(e- LMSs) lack ontologies for sharing their domain knowledge learning objects with others due to differences or non-uniformity in architectures, platforms, protocols and representations. The effect of this on e-learners is that collaboration with other e-LMS during learning processes is not permitted. Hence, learning process is restricted only to the knowledge base of a particular E-LMS adopted by an institution, which may limit the mastery level of learners. To provide a remedy to this problem, an intelligent multi-agent mediating system model is proposed in this study using hybrid rule and case based reasoning scheme. Unified Modeling Language(UML) is used as a design tool to specify the active and passive entities of the model in form class The model proposed provides a collaborative platform for sharing of the learning objects across multiple e-LMSs, during learning processes.
-
E-learning Materials Development: Applying and Implementing Software Reuse Principles and Granularity Levels in the SmallNabil Arman (SERSC, 2010-06-01)E-learning materials development is typically acknowledged as an expensive, complicated, and lengthy process, often producing materials that are of low quality and difficult to adaptand maintain. It has always been a challenge to identify proper e-learning materials that can be reused at a reasonable cost and effort. In this paper, software engineering reuse principlesare applied to e-learning materials development process. These principles are then applied and implemented in a prototype that is integrated with an open-source course management systems. The reuse of existing e-learning materials is beneficial in improving developers of elearning materials productivity. E-learning material reuse is performed, in this research, based on construct’s granularity rather than on unified constructs of one size.