The Visual Token Codec compresses ViT intermediate features by entropy-coding patch tokens on their native grid instead of a flattened sequence, reducing bitrate by 15.7x to 37.4x at 90% of uncompressed performance over a VTM-based baseline.
Vision transformers need registers,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Visual Token Codec: Unleashing Spatial Redundancy for ViT Feature Coding
The Visual Token Codec compresses ViT intermediate features by entropy-coding patch tokens on their native grid instead of a flattened sequence, reducing bitrate by 15.7x to 37.4x at 90% of uncompressed performance over a VTM-based baseline.