About this document
Multimodal Network for Paper-Cut Classification by ziyaogao99 is a document available to read on EtoBox.
The paper introduces a Multimodal Prototype Fusion Network (MPFN) for classifying paper-cut images, addressing challenges like artistic abstraction and data imbalance. Two variants, AMPFN and IMPFN, utilize CLIP for feature extraction, achieving high accuracy on seen and unseen classes, respectively. This approach enhances cultural heritage digitization and multimodal art analysis by providing a scalable solution for effective image classification.
- Author
- ziyaogao99
- Language
- EN