Pseudo-relevantfeedback概述.pptVIP

* In this work we mine the cognition of different communities using image search engine. Currently the successful image search engine employ the well know PageRank algorithm, which utilize the linking structure of the data. * . * * * Parameter Sensibility Parameter Sensibility Video Search Re-Ranking via Multi-Graph Propagation Contributions: Multi-modal fusion Employ an object-sensitive approach to query analysis to improve the baseline result of text-based video search. Propose a PageRank-like graph-based approach to text-based search result re-ranking. textual relevancy, semantic concept relevancy, low-level-feature-based visual similarity. Video Search Re-Ranking via Multi-Graph Propagation 注重于多模态融合 Video Search Re-Ranking via Multi-Graph Propagation Haveliwala proposed a topic-sensitive PageRank approach biased towards a set of representative topics, to capture more accurately the notion of importance with respect to a particular topic. In the topic-sensitive PageRank, the more relevance to the given topic a web page has, the more important the web page is. Video Search Re-Ranking via Multi-Graph Propagation A similar relative relevance dependent on the given topic also exists in the video search tasks. In video corpus, each video clip is annotated with a set of semantic concepts, which represent the semantic content of the video clip Video shots are not independent of each other, but have mutual relations such as conceptual and visual similarity. This can be taken as the underlying “hyperlink” between video shots, similar to that between web pages PageRank方法示意 * Page linking example 2. The images with higher PR values attract more attention The images with higher PR values are more valuable Query Analysis Text-based search baseline: Use the search results as the initialization of the graphs. N-gram Query Segmentation Segment the query into term sequences based on N-gram method Given a query like “find shots of one or more people reading a new