News

In a World where people want ever more access to video-based information, with the possibility of automatic visual search, automatic visual analysis, instant interaction and storing of huge amounts of visual data, the catch words for the user are “faster”, “easier” and “cheaper”.

The authors, with their work “Studying the Compression Performance of Video Descriptors”, goes straight to this point! Traditionally visual information has been represented using picture elements (pixels). However, there are challenging tasks related to computer vision applications that demand alternative representations. Visual feature descriptors are a powerful class of such representations. In scenarios strictly constrained in terms of storage capability, memory and network resources such as those observed in visual sensor networks and mobile visual search applications, compression of the description data may be imperative. When limited bandwidth is available, the description data has to be coded in efficient ways.

The main novelty of this work is the detailed benchmarking of the various coding modes for different types of video content in order to identify the best codec in terms of combination of modes. This work won the Best Paper Award in the SBrT 2015 – Brazilian Symposium of Telecommunications, which took place in Juiz de Fora."