Home › Categories › Technology› I2CDDE Algorithm

I2CDDE Algorithm.pdf

Accepted_manuscript.pdf

🔗 Source: ueaeprints.uea.ac.uk

📊 Size: 819 KB

👤 Author: Constance Roberts-Croot (RIN - Staff)

⬇️ Downloads: 152

Summary

A novel fully supervised local descriptor learning algorithm, called Image-to-Class Distances based Discriminative Embedding (I2CDDE), is proposed to learn compact but highly discriminative local feature descriptors for human action recognition. I2CDDE incorporates class labels and leverages the advantages of the image-to-class distance to enable fully supervised learning of local feature descriptors, achieving highly discriminative but compact local descriptors. The objective is to minimize the image-to-class distances from samples to their corresponding classes while maximizing the distances to other classes in the low-dimensional space. A manifold regularization based on the graph Laplacian is also incorporated to enhance the smoothness of the embedding. I2CDDE is applied to human action recognition on four benchmark datasets, achieving state-of-the-art performance and significantly improving I2C-based classifiers.

Download File

Description

A novel fully supervised local descriptor learning algorithm, called Image-to-Class Distances based Discriminative Embedding (I2CDDE), is proposed to learn...

Technical Information

File Format: PDF
File Size: 819 KB
Pages: 13
Language: EN
Author: Constance Roberts-Croot (RIN - Staff)
Total Downloads: 152
Last Updated: 4 weeks ago

Document Overview

This PDF document about I2CDDE Algorithm provides comprehensive information and guidance. Whether you're a beginner or advanced user, this resource offers valuable insights into I2CDDE Algorithm.