FocusVTC: Efficient and High-Performance Visual Text Compression with Adaptive Resolution

Open paper

Benchmark	Model	Rank	Results
optical-character-recognition-ocr-on-ocrbench	FocusVTC	#31	Score: 860
chart-question-answering-on-chartqa	FocusVTC	#35	1:1 Accuracy: 86.28
multimodal-reasoning-on-mmmu	FocusVTC	#46	Accuracy: 66.73