Cool-chic video: Learned video coding with 800 parameters

1 code implementation • 5 Feb 2024 • Thomas Leguay, Théo Ladune, Pierrick Philippe, Olivier Déforges

We propose a lightweight learned video codec with 900 multiplications per decoded pixel and 800 parameters overall.

Cool-Chic: Perceptually Tuned Low Complexity Overfitted Image Coder

1 code implementation • 4 Jan 2024 • Théo Ladune, Pierrick Philippe, Gordon Clare, Félix Henry, Thomas Leguay

This paper summarises the design of the Cool-Chic candidate for the Challenge on Learned Image Compression.

Image Compression

ED: Perceptually tuned Enhanced Compression Model

no code implementations • 4 Jan 2024 • Pierrick Philippe, Théo Ladune, Stéphane Davenet, Thomas Leguay

This paper summarises the design of the candidate ED for the Challenge on Learned Image Compression 2024.

Image Compression

Low-complexity Overfitted Neural Image Codec

1 code implementation • 24 Jul 2023 • Thomas Leguay, Théo Ladune, Pierrick Philippe, Gordon Clare, Félix Henry

We propose a neural image codec at reduced complexity which overfits the decoder parameters to each input image.

Decoder

COOL-CHIC: Coordinate-based Low Complexity Hierarchical Image Codec

1 code implementation • ICCV 2023 • Théo Ladune, Pierrick Philippe, Félix Henry, Gordon Clare, Thomas Leguay

At the receiver side, the compressed image is obtained by evaluating the mapping function for all pixel coordinates.