InterDigital

CompressAI-Vision helps you design, test and compare Video Compression for Machines pipelines. Compression methods can be either pulled from custom AI-based modules from CompressAI or traditional codecs such as H.266/VVC.

A Python Library for the 3GPP physical layer

Read and create MPEG Avatar Representation Format content.

A PyTorch library and evaluation platform for end-to-end compression research

Software model of hardware-friendly film grain synthesis, supporting MPEG FGC SEI and AOM/AV1 parameters

Deleting remote inference delay via on-device correction

Implementation of "CAFe-GS: Compactness-Aware Frequency-Guided Densification for 3D Gaussian Splatting", ICPR 2026

fork of VTM VVC Software from Fraunhofer HHI: https://vcgit.hhi.fraunhofer.de/jvet/VVCSoftware_VTM

Gen AI Traffic Modelling Framework