Unifying Keywords and Visual Contents in Image Retrieval
IEEE MultiMedia
Automatic Annotation and Retrieval of Images
World Wide Web
Proceedings of the 26th annual international ACM SIGIR conference on Research and development in informaion retrieval
Evaluating the impact of selection noise in community-based web search
Proceedings of the 28th annual international ACM SIGIR conference on Research and development in information retrieval
Image annotations by combining multiple evidence & wordNet
Proceedings of the 13th annual ACM international conference on Multimedia
A survey of content-based image retrieval with high-level semantics
Pattern Recognition
Computer Vision and Image Understanding
The Google Similarity Distance
IEEE Transactions on Knowledge and Data Engineering
Supervised Learning of Semantic Classes for Image Annotation and Retrieval
IEEE Transactions on Pattern Analysis and Machine Intelligence
Multiple Bernoulli relevance models for image and video annotation
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
| Hi-index | 12.05 |
State-of-the-art approaches for the semantic characterization of the visual content rely on segmentation techniques outputting static entities such as rectangular regions, blobs or multimedia objects. Once these entities are highlighted, additional external information allowing to refine this characterization within knowledge bases (e.g. the fact that semantic concepts sky and sun corresponding to these entities often co-occur) is ignored due to the difficulty of integrating it a priori within these static approaches. We formulate the hypothesis that this information is important in the process of highlighting the semantic visual content and propose an architecture based on image agents, abstract structures representing the visual entities of the image content, which integrate it dynamically. For this, we investigate the formation of semantic concepts within a population of multimedia agents. We first propose a learning framework linking the automatically-extracted visual content to these agents. We then develop an architecture allowing the communication of image agents about the perceived semantic concepts by taking into account the external information incrementally. We validate our proposition in a system-based evaluation framework on a corpus of real world color photographs.