Automatic Rating and Selection of Digital Photographs.
ABSTRACT This paper presents methods for automatically rating and selecting digital photographs. The importance of each photograph is es- timated by analyzing its content as well as its time-metadata. The pres- ence of people is estimated by combining face and skin detection. Finally, the appeal of each photograph is calculated using a trained SVM classi- fier. The results of a conducted user study show that the automatically obtained rating coincides well with the perception of the test persons.
- SourceAvailable from: Sabine Thieme
Conference Proceeding: MetaXa - Context- and Content-Driven Metadata Enhancement for Personal Photo Books.[show abstract] [hide abstract]
ABSTRACT: Making a photo book as a special gift to your beloved can be very time-consuming. One has to carefully select and arrange the pic- tures nicely over the pages of a previously bought photo book. In these days, photo finisher companies are able to directly print and bind a nice photo book from a selected set of images. But for the users of the soft- ware that comes with the album creation, the selection and arrangement of pictures in the album still remains a tedious task. What is missing are easy and good suggestions which pictures to select and how to arrange them into a personal photo book. A higher availability of metadata with the pictures could enable a content-driven and context-driven selection and make album creation better and easier. With MetaXa, we propose a flexible, component-based software architecture that iteratively allows for the multimodal extraction and enhancement of metadata for personal media content. The enhancement process is realized by extraction and enhancement components that each contribute to a well-defined annota- tion task. Depending on the application domain dierentAdvances in Multimedia Modeling, 13th International Multimedia Modeling Conference, MMM 2007, Singapore, January 9-12, 2007. Proceedings, Part I; 01/2007
- [show abstract] [hide abstract]
ABSTRACT: In this paper, algorithms for automatic albuming of consumer photographs are described. Specifically, two core algorithms namely event clustering and screening of low-quality images, are introduced and their performance is evaluated. Event clustering and image quality screening have many applications including albuming services, image management and organization, and digital photofinishing. These are difficult tasks because there is, in general, none (or very limited) contextual information about picture content, and the final interpretation could be subjective. A novel event-clustering algorithm is created to automatically segment pictures into events and subevents for albuming, based on date/time metadata information, as well as color content of the pictures. A block-based color histogram correlation technique is developed for image content comparison of general consumer pictures. A new quality-screening algorithm is developed based on object quality measures, to detect problematic images caused by underexposure, low contrast, and camera defocus or movement.IEEE Transactions on Multimedia 10/2003; · 1.75 Impact Factor
Conference Proceeding: Time as essence for photo browsing through personal digital libraries.[show abstract] [hide abstract]
ABSTRACT: We developed two photo browsers for collections with thousands of time-stamped digital images. Modern digital cameras record photo shoot times, and semantically related photos tend to occur in bursts. Our browsers exploit the timing information to structure the collections and to automatically generate meaningful summaries. The browsers differ in how users navigate and view the structured collections. We conducted user studies to compare the two browsers and an un-summarized image browser. Our results show that exploiting the time dimension and appropriately summarizing collections can lead to significant improvements. For example, for one task category, one of our browsers enabled a 33% improvement in speed of finding given images compared to the commercial browser. Similarly, users were able to complete 29% more tasks when using this same browser.ACM/IEEE Joint Conference on Digital Libraries, JCDL 2002, Portland, Oregon, USA, June 14-18, 2002, Proceedings; 01/2002