The visual quality assessment approaches and their classification are introduced. Recent developments on both image and video quality metrics, as well as several publicly available databases for both images and videos, are reviewed. Then, we conduct experiments on image and video quality databases to compare the performance of some existing state-ofthe-art visual quality metrics. It is shown th...

In this paper we describe new methods to detect semantic concepts from digital video based on audible and visual content. Temporal Gradient Correlogram captures temporal correlations of gradient edge directions from sampled shot frames. Power-related physical features are extracted from short audio samples in video shots. Video shots containing people, cityscape, landscape, speech or instrument...

Video-assisted thoracic surgery has proven to be safe and effective for the diagnosis and management of pericardial disease. Three-dimensional (3-D) video imaging technology has been developed to allow the laparoscopic surgeon more precision and efficiency in advanced laparoscopic cases. This case report describes the marriage of 3-D video imaging and thoracoscopy that allowed performance of a ...

We propose a probabilistic video model, the Video Pixel Network (VPN), that estimates the discrete joint distribution of the raw pixel values in a video. The model and the neural architecture reflect the time, space and color structure of video tensors and encode it as a fourdimensional dependency chain. The VPN approaches the best possible performance on the Moving MNIST benchmark, a leap over...

Watermarking is an advanced technology that identifies to solve the problem of illegal manipulation and distribution of digital data. It is the art of hiding the copyright information into host such that the embedded data is imperceptible. The covers in the forms of digital multimedia object, namely image, audio and video. The extensive literature collected related to the performance improvemen...

This paper gives an overview of state of the art in algorithms and tools using video and audio features. In particular, we extensively survey the features that can be efficiently extracted from compressed video and audio, which include visual, motion, coding, and audio features. We also discuss typical applications of these compressed-domain features in video and audio indexing and analysis. Fi...

We propose a framework for temporally consistent video completion. To this end we generalize the exemplar-based inpainting method of Criminisi et al . [7] to video inpainting. Specifically we address two important issues: Firstly, we propose a color and optical flow inpainting to ensure temporal consistency of inpainting even for complex motion of foreground and background. Secondly, rather tha...

In large video collections with clusters of typical categories, such as “birthday party” or “flash-mob”, category-specific video summarization can produce higher quality video summaries than unsupervised approaches that are blind to the video category. Given a video from a known category, our approach first efficiently performs a temporal segmentation into semantically-consistent segments, deli...

Always-on media spaces broadcast video between collaborators to provide mutual awareness and to encourage casual interaction. This video can be easily recorded on the fly as a video trace. Ostensibly, people can review this video history to gain a better idea of the activities and availability of their collaborators. Such systems are obviously highly contentious, as they raise significant priva...

Online video advertising (video-in-video) strategies are typically agnostic to the video content (ex.: advertising on YouTube) and the human viewer’s preferences. How to assess the emotional state and engagement of the viewer to place an advertisement? Where to insert an advertisement based on the content in an advertisement and a specific target video stream? Surely these are relevant question...