Visualisation and Browsing of Flickr Retrieval Results

With the prevalence of digital cameras and increased band with, online photo collections are growing at a rapid rate. Especially larger image collections require user-friendly tools to search and browse the dataset for which standard approaches of organising images are not sufficient. In this demo paper, we present an efficient and effective approach to visualising and browsing image search results. While various approaches to visualisation and browsing of image database have been introduced, these are too slow to be applied to online retrieval results since the complete images would need to be downloaded for feature extraction and exploitation. The method we present here addresses this problem and allows for interactive image database navigation of images retrieved from the photo sharing site Flickr. We do this by utilising solely the headers of JPEG files, in particular the Huffman tables there, for feature extraction coupled with a projection onto a 2-dimensional visualisation space based on principal component analysis. Images are dynamically placed into a grid structure and organised in a tree-like hierarchy for visual browsing.