GIViC introduces a novel generative implicit framework for video compression that advances the performance limits of neural representation-based codecs.

• Leverages implicit neural representations optimized with a transformer architecture
• Achieves state-of-the-art performance compared to conventional compression methods
• Demonstrates the potential of generative AI approaches to transform video compression technology
• Reduces bandwidth requirements while maintaining high visual quality

This research represents a significant engineering breakthrough in how video content can be compressed and transmitted efficiently—critical for streaming services, AR/VR applications, and video conferencing platforms facing bandwidth constraints.

GIViC: Generative Implicit Video Compression