Material Links
MEDIA ANALYTICS
Link to: MA Projects
PROJECTS
Link to: MA People
PEOPLE
Link to: MA Publications
PUBLICATIONS
Link to: MA Patents
PATENTS
Embedding Label Structures for Fine-Grained Feature Representation
Follow a manual added link
CVPR 2016 | We model the multi-level relevance among fine-grained classes for fine-grained categorization. We jointly optimize classification and similarity constraints in a proposed multi-task learning framework, and we embed label structures such as hierarchy or shared attributes into the framework by generalizing the triplet loss. This significantly outperforms previous fine-grained feature representations for image retrieval at different levels of relevance.
Collaborators: Xiaofan Zhang, Feng Zhou, Yuanqing Lin, Shaoting Zhang
Media Analytics News
Nothing Found
Sorry, no posts matched your criteria