Multi-Instance Learning Based Approach to Image Retrieval
Wei Kun-juan · Zhongwen xinxi xuebao · 2008
Multi-instance learning has been employed in Content-Based Image Retrieval(CBIR) for it's gracefully performance in solving the ambiguity of image.The whole image is regarded as a multi-instance bag.The image is partitioned into several semantic regions by using an approach—based on gauss mixing model and improved EM clustering to segment the image,then the regions described by color,texture,shape,and invariant moment features are regarded as the instances in the bag.Next,query images posed by the user are transformed into corresponding positive and negative bags and a multi-instance algorithm is employed for image retrieval and relevance feedback.Experiments show this approach get a better result than the other methods.