Conventional video summarization methods focus predominantly on summarizing videos along the time axis, such as building a movie trailer. The resulting video trailer tends to reta...
This paper 1 addresses the problem of efficient visual 2D template tracking in image sequences. We adopt a discriminative approach in which the observations at each frame yield di...
We address the problem of visual category recognition by learning an image-to-image distance function that attempts to satisfy the following property: the distance between images ...
Andrea Frome, Yoram Singer, Fei Sha, Jitendra Mali...
In this paper a fully automatic system for embedding visually recognizable watermark patterns to video objects is proposed. The architecture consists of 3 main modules. During the...
Nikolaos D. Doulamis, Anastasios D. Doulamis, Stef...
The Web has become an excellent source for gathering consumer opinions. There are now numerous Web sites containing such opinions, e.g., customer reviews of products, forums, disc...