Title :
Active learning and its scalability for image retrieval
Author :
Chang, Edward Y. ; Lai, Wei-Cheng
Author_Institution :
California Univ., Santa Barbara, CA, USA
Abstract :
Active learning has been shown to be a viable tool for learning complex, subjective query concepts with a small number of training instances. In this work, we compare four active-learning algorithms and study the best sample-selection strategies. We also discuss two scalability issues of active learning: scalability in dataset size, and scalability in concept complexity. To address these challenges, we suggest future directions that research might take.
Keywords :
image retrieval; learning (artificial intelligence); query formulation; support vector machines; SVM; active learning; angle-diversity algorithm; concept complexity scalability; concept diversity; concept isolation; concept scarcity; dataset size scalability; error-reduction algorithm; image retrieval; kernel methods; sample-selection strategies; speculative active learning; subjective query concepts; support vectors; target query-concept formulation; training instances; Binary trees; Feedback; Image retrieval; Kernel; Performance analysis; Performance evaluation; Scalability; Support vector machine classification; Support vector machines;
Conference_Titel :
Multimedia and Expo, 2004. ICME '04. 2004 IEEE International Conference on
Print_ISBN :
0-7803-8603-5
DOI :
10.1109/ICME.2004.1394128
