In this paper, we propose a novel language-model-based approach with pseudo-relevance feedback for addressing the vocabulary problem in visual BOW mining. We ...

This paper proposes a novel language-model-based approach with pseudo-relevance feedback for addressing the vocabulary problem in visual BOW mining, ...

One leading framework for image object mining is the bag-of-words (BOW) approach. The idea is to encode an image as a collection of visual words of the ...

In this paper, we propose a novel language-model-based approach with pseudo-relevance feedback for addressing the vocabulary problem in visual BOW mining. We ...

A Novel Language-Model-Based Approach for Image Object Mining and Re-Ranking. Authors: Hsiao, Jen-Hao · Chen, Chu-Song · Chen, Ming-Syan. Issue Date: 2008-12-01.

A Novel Language-Model-Based Approach for Image Object Mining and Re-ranking ; Near-duplicate keyframe retrieval with visual keywords and semantic context. Wu X.

Experiments show that language model based re-ranking method is effective in presenting more relevant documents on the top retrieved results to mobile users.

標題: A novel language-model-based approach for image object mining and re-ranking ; 作者: Hsiao, J.-H. · Chen, C.-S. · MING-SYAN CHEN ; 公開日期: 2008 ; 起(迄)頁: ...

In this paper, we propose a new bag based re ranking framework for large scale TBIR. We compute this problem as Multiple Instance Learning and Generalized ...

Dec 15, 2008 · One leading framework for image object mining is the bag-of-words (BOW) approach. The idea is to encode an image as a collection of visual words ...