Evaluation Methodology for Image Testbed and Content-Based Retrieval
Aidong Zhang Principal Investigator
MetadataShow full item record
The project aims at developing an evaluation methodology for content-based <br/>image retrieval. The research involves: (1) Measurement of the complexity of <br/>image testbeds which can be used to quantitatively determine the degree of <br/>difficulty in retrieving images from the image testbeds, and (2) Comparison <br/>of the performance of different retrieval approaches which can quantitatively <br/>give an objective ranking of the performance of the retrieval approaches. <br/>The project designs a general framework of image feature representations <br/>which can be used as a vehicle to conduct statistical analysis on images and <br/>forms a basis for establishing the evaluation methodology. Using this framework,<br/>the evaluation method can measure the complexity of the image databases by <br/>their cross entropy and rank the retrieval approaches by their cross entropy <br/>with respect to a particular testbed. The image testbeds can thus be compared <br/>with each other on their complexity in supporting image querying. Furthermore, <br/>the retrieval techniques can be compared with each other without using queries <br/>so the human subjectivity is avoided. The algorithms developed in this project <br/>will be a valuable asset for the community to eventually establish a general <br/>theory of the evaluation methodology for content-based image retrieval research.