Multi-task convolution neural network-based lifting scheme for image compression

Tassnim Dardouri

doi:10.1016/j.patrec.2025.05.001

Multi-task convolution neural network-based lifting scheme for image compression

2025-05-20

0

OA

AI

Amel Benazza‐Benyahia

G

Gabriel Dauphin

DOI：10.1016/j.patrec.2025.05.001

Original

Original request for help

Save

Abstract

En 中文

• Investigating neural networks for the design of lifting-based image coding systems. • Building a CNN-based lifting architecture to better capture the local structure of the image to be encoded. • Developing a multi-task learning approach exploiting the similarities between prediction stages. • Illustrating the benefits of the proposed approach compared to recent related works through objective and subjective evaluation.

Multi-task convolution neural network-based lifting scheme for image compression

Abstract

Journal

Organization