A text-prompted ROI image coder with a user-controlled quality knob and latent mask attention achieves strong RD and machine-vision results, though headline numbers use ground-truth masks.
Just noticeable difference for deep machine vision,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Customizable ROI-Based Deep Image Compression
A text-prompted ROI image coder with a user-controlled quality knob and latent mask attention achieves strong RD and machine-vision results, though headline numbers use ground-truth masks.