TokBench measures text recognition accuracy and face similarity on reconstructed images and videos across 16 tokenizers, showing small text and faces are poorly preserved and often missed by traditional metrics.
Imagenet: A large- scale hierarchical image database
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
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
TokBench: Evaluating Your Visual Tokenizer before Visual Generation
TokBench measures text recognition accuracy and face similarity on reconstructed images and videos across 16 tokenizers, showing small text and faces are poorly preserved and often missed by traditional metrics.