Defesa de Mestrado de Edward Cayllahua Cashuina; dia 22/02/2013, as 10:00, Sala de Seminários.

Defesa de Mestrado de Edward Cayllahua Cashuina; dia 22/02/2013, as 10:00, Sala de Seminários.

Título: Static Video Summarization Based on Local Descriptors and Temporal Segmentation.

Resumo: During the last years, a continuous demand and creation of digitalvideo information have occurred. The creation of digital video has caused an exponential growth of digital video content. To increase the usability of such large volume of videos, a lot of research has been made. Video summarization, in particular, has been proposed to rapidly browse large video collections. It has also been used to e ciently index and access video content. To summarize any type of video, researchers have relied on visual features contained in frames. In order to extract these features,     different techniques have used local or global descriptors. Nonetheless, no extensive evaluation have been made about the usefulness of both types of descriptors in video summarization. In this dissertation, we perform a wide evaluation in order to achieve a stronger position about the performance of local descriptors in semantic video summarization. Furthermore, we evaluate how local descriptors are a ected by video segmentation and we also evaluate the robustness of local descriptors compared to global descriptors. Additionally, we inspect if color information can help local descriptors to produce better video summaries. Another important contribution of this dissertation is to propose a simple model for semantic video summarization that can produce meaningful and informative video summaries.