Title :
A novel semi-supervised Multi-Instance learning approach for scene recognition
Author :
Li Jun-yi ; Li Jian-hua
Author_Institution :
Sch. of Electron. Inf. & Electr. Eng., Shanghai Jiao Tong Univ., Shanghai, China
Abstract :
We proposes a new image Multi-Instance (MI) bag generating method, which models an image with a Gaussian Mixed Model (GMM). The generated GMM is treated as an MI bag, of which the color and locally stable invariant components (SIFT) are the instances. Agglomerative Information Bottleneck clustering is employed to transform the MIL problem into single-instance learning problem so that single-instance classifiers can be used for classification. Finally, ensemble learning is involved to further enhance classifiers´ generalization ability. Experimental results demonstrate that the performance of the proposed framework for image recognition is superior to some common MI algorithms on average in a 5-category scene recognition task.
Keywords :
Gaussian processes; image recognition; learning (artificial intelligence); 5-category scene recognition task; Gaussian mixed model; SIFT; agglomerative information bottleneck clustering; ensemble learning; image multiinstance bag generating method; image recognition; semisupervised multiinstance learning; Accuracy; Classification algorithms; Clustering algorithms; Feature extraction; Image color analysis; Support vector machine classification; Training; AIB Clustering; Ensemble Classifier; Gaussian Mixed Model; Multi-Instance Learning; Scene Recognition; Single-Instance Bag; image modeling;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery (FSKD), 2012 9th International Conference on
Conference_Location :
Sichuan
Print_ISBN :
978-1-4673-0025-4
DOI :
10.1109/FSKD.2012.6234079