Hybrid Spatial–Spectral Autoencoder Models for Lossy Satellite Image Compression

What is it about?

This paper introduces advanced deep learning-based hybrid autoencoder models, incorporating Convolutional Neural Networks, Long Short-Term Memory, and Attention mechanisms for efficient satellite image compression, outperforming JPEG and state-of-the-art models in quality metrics and demonstrating compatibility with AI-based chipsets. A case study conducted on a real satellite scene demonstrates the compatibility of the proposed models with recent AI-based chipsets.

Featured Image

Why is it important?

This paper introduces advanced deep learning-based hybrid autoencoder models, incorporating Convolutional Neural Networks, Long Short-Term Memory, and Attention mechanisms for efficient satellite image compression, outperforming JPEG and state-of-the-art models in quality metrics and demonstrating compatibility with AI-based chipsets. A case study conducted on a real satellite scene demonstrates the compatibility of the proposed models with recent AI-based chipsets.

Perspectives