1. Mining Weakly Labeled Web Facial
Images for Search-Based Face
Annotation
2. ABSTRACT:
Auto face annotation is playing vital role in many real-world
knowledge management systems and multimedia information.
To solve general content based face annotation problem, The
ULR is used for refining the web facial images.
Clustering based approximation improve the performance of
search based face annotation scheme.
4. Search Based Face
Annotation:
ULR algorithm To enhance the label quality.
Clustering-based approximation solution To
improve the scalability
5. Unsupervised Label refinement
To refine the web facial images using search based where face image
as query image.
The unsupervised label refinement algorithm which enhanced new
label matrix
6. Clustering based approximation
To improve the performance of search based face annotation
scheme.
Images are grouped into clusters for refining process more easier.
It provides high performance with clusters instead of refining
single image annotation