In this paper, we study the problem of Web forum crawling. Web forum has now become an important data source of many Web applications; while forum crawling is still a challenging task due to complex in-site link structures and login controls of most forum sites. Without carefully selecting the traversal path, a generic crawler usually downloads many duplicate and invalid pages from forums, and thus wastes both the precious bandwidth and the limited storage space. To crawl forum data more effectively and efficiently, in this paper, we propose an automatic approach to exploring an appropriate traversal strategy to direct the crawling of a given target forum. In detail, the traversal strategy consists of the identification of the skeleton links and the detection of the page-flipping links . The skeleton links instruct the crawler to only crawl valuable pages and meanwhile avoid duplicate and uninformative ones; and the page-flipping links tell the crawler how to completely download a long discussion thread which is usually shown in multiple pages in Web forums. The extensive experimental results on several forums show encouraging performance of our approach. Following the discovered traversal strategy, our forum crawler can archive more informative pages in comparison with previous related work and a commercial generic crawler.

We study in this paper the Web forum crawling problem, which is a very fundamental step in many Web applications, such as search engine and Web data mining. As a typical user-created content (UCC), Web forum has become an important resource on the Web due to its rich information contributed by millions of Internet users every day. However, Web forum crawling is not a trivial problem due to the in-depth link structures, the large amount of duplicate pages, as well as many invalid pages caused by login failure issues. In this paper, we propose and build a prototype of an intelligent forum crawler, iRobot, which has intelligence to understand the content and the structure of a forum site, and then decide how to choose traversal paths among different kinds of pages. To do this, we first randomly sample (download) a few pages from the target forum site, and introduce the page content layout as the characteristics to group those pre-sampled pages and re-construct the forum's sitemap. After that, we select an optimal crawling path which only traverses informative pages and skips invalid and duplicate ones. The extensive experimental results on several forums show the performance of our system in the following aspects: 1) Effectiveness - Compared to a generic crawler, iRobot significantly decreases the duplicate and invalid pages; 2) Efficiency - With a small cost of pre-sampling a few pages for learning the necessary knowledge, iRobot saves substantial network bandwidth and storage as it only fetches informative pages from a forum site; and 3) Long threads that are divided into multiple pages can be re-concatenated and archived as a whole thread, which is of great help for further indexing and data mining.

In this paper, we proposed a novel prototype of automated music video generation using web image resource. In this prototype, the salient words/phrases of a song's lyrics are first automatically extracted and then used as queries to retrieve related high-quality images from web search engines. To guarantee the coherence among the chosen images' visual representation and the music song, the returned images are further re-ranked and filtered based on their content characteristics such as color, face, landscape, as well as the song's mood type. Finally, those selected images are concatenated to generate a music video using the Photo2Video technique, based on the rhythm information of the music. Preliminary evaluations of the proposed prototype have shown promising results.

This paper1 is concerned with the problem of multimodal fusion in video search. First, we employ an object-sensitive approach to query analysis to improve the baseline result of text-based video search. Then, we propose a PageRank-like graph-based approach to text-based search result re-ranking. To better exploit the underlying relationship between video shots, the proposed re-ranking scheme simultaneously leverages textual relevancy, semantic concept relevancy, and low-level-feature-based visual similarity. In this PageRank-like scheme, we construct a set of graphs with the video shots as vertexes, and the conceptual and visual similarity between video shots as "hyperlinks". A modified topic-sensitive PageRank algorithm is then applied on these graphs to propagate the relevance scores through all related video shots. Experimental results verify the effectiveness of the graph-based propagation approach combined with the object-sensitive query analysis approach, which brings significant improvement to the baseline of text-based video search. Our experimental analysis also indicates that the proposed re-ranking method is highly generic and independent of different query classes, training data, and human interference.

With the rapid development of telecommunication techniques and digital devices, it is quite easy to copy, modify and republish videos in digital format, resulting in large volume of duplicate videos on the Web in recent years. In this paper we mainly investigate the problem of detecting excessive content duplication, so as to facilitate video search and intelligence propriety protection. A real-time detection method is hence proposed, which first selects videos' representative frames and then reduces each to a 64 bit hash code. Then the similarity of any two videos can be estimated by the proportion of their similar hash codes. The experiments demonstrate that our approach is both efficient and effective in terms of real-time applications.

This paper describes the MSRA-USTC-SJTU experiments for TRECVID 2007. We performed the experiments in high-level feature extraction and automatic search tasks. For high-level feature extraction, we investigated the benefit of unlabeled data by semi-supervised learning, and the multi-layer (ML) multi-instance (MI) relation embedded in video by MLMI kernel, as well as the correlations between concepts by correlative multi-label learning. For automatic search, we fuse text, visual example, and concept-based models while using temporal consistency and face information for re-ranking and result refinement.

The explosive growth of information technology and digital content industry stimulates various video applications over the Internet. Since it is quite easy to copy, reformat, modify and republish video files on the websites, similarity/duplicate detection and measurement is essential to identify the excessive content duplication, so as to facilitate effective video search and intelligence propriety protection as well. In this paper, we propose a novel signature-based approach for duplicate video comparison. The so-called video histogram scheme counts the numbers of video’s frames that are closest to a set of representative seed vectors chosen from the feature space of the training data set in advance. Then all the numbers are normalized to generate the signature of the video for further comparison. As our signature is a compact fixed-size vector with low dimension for each video, it requires less storage and computation cost than previous methods. The experiments show that our approach is both efficient and effective for web video duplicate detection.

In this paper, we describe the MSRA experiments for TRECVID 2006, including details of the approaches and performance analyses for high-level feature extraction task and rushes exploitation task. For high-level feature extraction, we mainly investigated the benefit of unlabeled data by semi-supervised learning methods, including adaptive semi-supervised learning with kernel density estimation, manifold ranking, and transductive graph. Moreover, we performed fusion in two different levels: modality level and model level. We were ranked in the top 10 list in terms of mean average precision performance among all participants. For rushes exploitation, we detected the duplicate content based on ordinal video signature. We also performed video structuring (i.e. decomposing rushes into shots and sub-shots) and camera motion classification (i.e. classifying each sub-shot into static, pan, tilt, zoom, rotation, or object motion in terms of camera motion). Furthermore, we validated the approaches to concept modeling and detected 39 concepts on rushes data without re-training the visual models obtained in high-level feature extraction task.

Most existing web video search engines index videos by file names, URLs, and surrounding texts. These types of video metadata roughly describe the whole video in an abstract level without taking the rich content, such as semantic content descriptions and speech within the video, into consideration. Therefore the relevance ranking of the video search results is not satisfactory as the details of video contents are ignored. In this paper we propose a novel relevance ranking approach for Web-based video search using both video metadata and the rich content contained in the videos. To leverage real content into ranking, the videos are segmented into shots, which are smaller and more semantic-meaningful retrievable units, and then more detailed information of video content such as semantic descriptions and speech of each shots are used to improve the retrieval and ranking performance. With video metadata and content information of shots, we developed an integrated ranking approach, which achieves improved ranking performance. We also introduce machine learning into the ranking system, and compare them with IR-model (information retrieval model) based method. The evaluation results demonstrate the effectiveness of the proposed ranking methods.

This paper presents an efficient learning scheme for automatic annotation of video shot size. Instead of existing methods that applied in sports videos using domain knowledge, we are aiming at a general approach to deal with more video genres, by using a more general low- and mid- level feature set. Support Vector Machine (SVM) is adopted in the classification task, and an efficient co-training scheme is used to explore the information embedded in unlabeled data based on two complementary feature sets. Moreover, the subjectivity-consistent costs for different mis-classifications are introduced to make the final decisions by a cost minimization criterion. Experimental results indicate the effectiveness and efficiency of the proposed scheme for shot size annotation.

This paper presents an automatic video genre categorization scheme based on the hierarchical ontology on video genres. Ten computable spatio-temporal features are extracted to distinguish the different genres using a hierarchical support vector machines (SVM) classifier built by cross-validation, which consists of a series of SVM classifiers united in a binary-tree form. As the order and genre partition strategy of the SVM classifier series affect the over performance of the united classifier, two optimal SVM binary trees, local and global, are constructed aiming at finding the best categorization orders, i.e., the best tree structure, of the genre ontology. Experimental results show that the proposed scheme outperforms C4.5 decision tree, typical 1-vs-1 SVM scheme, as well as the hierarchical SVM built by K-means.

Traditional video coding and streaming algorithms have not been fully utilized by using video content information for bit allocation and adaptation. This paper proposes a content-based video streaming method, based on a visual attention model to better utilize network bandwidth and achieve better subjective video quality. First, the visual attention model is exploited to segment the regions of interest (ROI) in video frames. Then, considering that the ROI is more sensitive to coding error than other regions, a region-weighted rate-distortion model is developed to allocate suitable bits for all ROI and non-ROI regions. The evaluation results indicate that more than 60% of the test video sequences encoded by the proposed method can obtain better subjective visual quality compared to the video encoded by classical methods under the same bandwidth, and about 20% of them can not be distinguished.

To maximize user satisfaction of video streaming, there is a tradeoff between spatial quality (image clarity) and temporal quality (motion smoothness) under a limited bandwidth. How to balance the requirement for the two aspects is a subjective selection. In this paper, we firstly introduce how to efficiently measure the subjective perceived spatial and temporal quality. Then we present a visual perception model to predict viewer’s satisfaction given the perceived spatial quality and temporal quality based on the result of a user study. At last, an adaptive video streaming system utilizing the perception model is proposed, which can automatically choose dropping fames or cutting the scalable bitstream according to the variable bandwidth, and obtain maximized perceptual quality.

Conventional frame rate up-conversion adopted in low bit-rate video coding/streaming cannot obtain good quality due to artifacts caused by composition problems. In this paper, a region based multiple frame-rate tradeoff scheme is proposed to reduce artifacts. The video frames are divided into regions of interest (ROIs) and backgrounds by the visual attention model. The motion intensity and complexity of ROI and background are computed by perceptive motion energy spectrum (PMES) to decide the ratio of the frame rate of ROI to that of background. The background is encoded at a low frame rate, and up-converted to the same frame rate as ROI at the decoder, with an integrated background motion composition model. The experimental results indicate that, compared to conventional frame rate up-conversion and normal codec scheme without frame rate up-conversion at the same bandwidth, the proposed scheme has higher PNSR performance and better subjective visual quality.

In this paper, a HMM-based visual speech system driven by audio speech input is designed to render a face model while synchronous audio is played. Compared to many methods adopted by other researchers, there is much difference between our approach and theirs. We first train the models for every final and initial in mandarin. In this process, a large quantity of audio training data under different surroundings and spoken by different people are used. Then, the recorded synchronous audiovisual speech data are used to make the trained models more adaptive to our specific announcer. Such models are more robust in synthesis phase and satisfying performance can be achieved even when input audio speech is degraded by noises.