Generative Video Compression with One-Dimensional Latent Representation

Dimensions overall score 8.0

GitHub Code Pulse

No public code linked for this paper yet.

• GVC1D achieves bitrate reductions of 60.4% under LPIPS on the HEVC Class B dataset.result 95%
• GVC1D achieves bitrate reductions of 68.8% under DISTS on the HEVC Class B dataset.result 95%
• GVC1D surpasses previous video compression methods in compression efficiency.result 90%
• Traditional 2D latent grid approaches preserve intra-frame redundancy due to their rigid structure.limitation 85%
• 2D latent grids are less effective for modeling long-term correlations in a compact and semantically coherent manner.limitation 85%
• GVC1D encodes video data into extreme compact 1D latent tokens conditioned on both short- and long-term contexts.method 90%
• 1D latent tokens can adaptively attend to semantic regions and naturally facilitate token reduction, thereby reducing spatial redundancy.technical 85%
• The proposed 1D memory provides semantically rich long-term context while maintaining low computational cost, thereby further reducing temporal redundancy.technical 85%

Competitive landscape

Competitor map is still being generated for this paper. Enable generation or check back soon.

Startup potential card