Title :
Ensembling Neural Networks-based 3d model retrieval
Author :
Liu, Yujie ; Li, Zongmin ; Li, Hua
Author_Institution :
Sch. Of Comput. Sci. & Commun. Eng., China Univ. Of Pet., Dongying
Abstract :
How to retrieve more and more 3D models is the important point in pervasive computing. In this paper, a novel ensembling neural network (NN) - based 3D model retrieval method is proposed. Firstly, four NNs are trained by constructed learning algorithm (CLA). These four NNs are trained by using difference feature of models. And then NN assembles are employed to retrieve 3D models. The experiments show that the supervised learning increased the power of retrieval.
Keywords :
image retrieval; learning (artificial intelligence); neural nets; solid modelling; ubiquitous computing; 3D model retrieval; constructed learning algorithm; ensembling neural network; pervasive computing; supervised learning; Assembly; Content based retrieval; Feature extraction; Feedback; Linear discriminant analysis; Neural networks; Pervasive computing; Shape; Supervised learning; Support vector machines; 3D model retrieval; Ensembling; Neural Network; Pervasive Computing;
Conference_Titel :
Pervasive Computing and Applications, 2008. ICPCA 2008. Third International Conference on
Conference_Location :
Alexandria
Print_ISBN :
978-1-4244-2020-9
Electronic_ISBN :
978-1-4244-2021-6
DOI :
10.1109/ICPCA.2008.4783631
