Video search reranking via information bottleneck principle

By Winston H. Hsu, L. Kennedy, Shih-Fu Chang. ACM Multimedia, 2006

307 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/4d99a54a512ea1c9ef1ada32ae8c4d3242dcd86e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Content-based visual search learned from social media (2012, methodology): “To find the dominant visual patterns, density estimation methods are often used, typically in the form of clustering [46,90] and random walk [52].”
Easy Samples First: Self-paced Reranking for Zero-Example Multimedia Search (2014, methodology): “In Clustering-based reranking [9], the retrieved videos are aggregated into clusters, and the clusters’ conditional probabilities of the pseudo samples are used to obtain a reranked list.”
How flickr helps us make sense of the world: context and content in community-contributed media collections (2007, background): “While this approach can still be quite powerful, recent efforts have suggested that sampling pseudo-negative examples from outside of the initial candidate set and learning light-weight discriminative models can actually greatly improve the performance of image ranking for a number of applications [”
Click Prediction for Web Image Reranking Using Multimodal Sparse Coding (2014, methodology): “Hsu et al. [36] also adopt this pseudo-positive and pseudo-negative image method to develop a clustering-based re-ranking algorithm.”
Multimodal Fusion for Video Search Reranking (2010, methodology): “Therefore, it is more crucial to offer high accuracy on the top-ranked documents than to improve the whole search performance on the entire result list [5].”
An Attribute-Assisted Reranking Model for Web Image Search (2015, methodology): “To verify the effectiveness of the proposed attribute-assisted reranking method, we compare the following approaches for performance evaluation: • Information Bottle [9].”
Learning to Rank Using User Clicks and Visual Features for Image Retrieval (2015, background): “Another feasible solution for combining visual information is visual reranking [14], [15], [39], [40], which combines both the textual and visual information and returns visually satisfying retrieved results.”