A Machine Learning Approach to 3D Model Retrieval
A novel 3D model retrieval system based on the machine learning approach is proposed in this work. This approach can be integrated with any existing 3D model matching algorithm which includes 3D model feature extraction and distance computation. By calculating the variance of each component among all feature vectors and removing those components with a larger variance, we can reduce the feature dimension. Furthermore, we derive the SVM classifier based on features extracted from the training set. We conduct experiments using the McGill Articulated Shape Benchmark database 1 for 3D model classification and retrieval, and demonstrate a significant performance improvement in the precision-recall curves.
Kuan-Hsien Liu Pei-Ying Chiang C.-C. Jay Kuo
Ming Hsieh Department of Electrical Engineering and Signal and Image Processing Institute University Ming Hsieh Department of Electrical Engineering and Signal and Image Processing InstituteUniversity
国际会议
2011亚太信号与信息处理协会年度峰会(APSIPAASC 2011)
西安
英文
1-5
2011-10-18（万方平台首次上网日期，不代表论文的发表时间）