MR-MIL : Manifold Ranking Based Multiple-Instance Learning for Automatic Image Annotation
スポンサーリンク
概要
- 論文の詳細を見る
A novel automatic image annotation (AIA) scheme is proposed based on multiple-instance learning (MIL). For a given concept, manifold ranking (MR) is first employed to MIL (referred as MR-MIL) for effectively mining the positive instances (i. e. regions in images) embedded in the positive bags (i. e. images). With the mined positive instances, the semantic model of the concept is built by the probabilistic output of SVM classifier. The experimental results reveal that high annotation accuracy can be achieved at region-level.
- （社）電子情報通信学会の論文
- 2008-10-01
著者
-
ZHU Zhenfeng
Institute of Information Science, Beijing Jiaotong University
-
Zhao Yao
Institute Of Information Science Beijing Jiaotong University
-
Pan Jeng-shyang
Department Of Electronic Engineering Kaohsiung University Of Applied Sciences
-
Pan Jeng-shyang
Department Of Automatic Test And Control Harbin Institute Of Technology
-
Zhu Zhenfeng
Institute Of Information Science Beijing Jiaotong University
-
Zhu Zhenfeng
Beijing Jiaotong Univ. Beijing Chn
-
ZHAO Yufeng
Institute of Information Science, Beijing Jiaotong University
-
Zhao Yufeng
Institute Of Information Science Beijing Jiaotong University
関連論文
- Commercial Shot Classification Based on Multiple Features Combination
- A Digital Image Watermarking Method Based on Labeled Bisecting Clustering Algorithm (Information Security)
- A Digital Image Watermarking Method Based on Labeled Bisecting Clustering Algorithm
- Content Adaptive Visible Watermarking during Ordered Dithering(Application Information Security)
- Visible Watermarking for Halftone Images(Information Security)
- Video Watermarking by Space-Time Interest Points
- Lossless Data Hiding Based on Companding Technique and Difference Expansion of Triplets(Image)
- A Fast K Nearest Neighbors Classification Algorithm(Image)
- A Subsampling-Based Digital Image Watermarking Scheme Resistant to Permutation Attack
- Texture Segmentation Using Separable and Non-Separable Wavelet Frames (Special Section on Digital Signal Processing)
- Genetic Feature Selection for Texture Classification Using 2-D Non-separable Wavelet Bases(Special Section on Digital Signal Processing)
- Attacking Phase Shift Keying Based Watermarking(Image)
- Commercial Shot Classification Based on Multiple Features Combination
- High Capacity Watermark Embedding Based on Invariant Regions of Visual Saliency
- MR-MIL : Manifold Ranking Based Multiple-Instance Learning for Automatic Image Annotation
- Compatible Stereo Video Coding with Adaptive Prediction Structure
- Robust Multi-Bit Watermarking for Free-View Television Using Light Field Rendering