MMSD and SAMR achieve 99 percent and 99.1 percent average data reduction for traffic images by transmitting segmentation maps, edges, text or semantically masked JPEGs and reconstructing via diffusion or inpainting models.
SQ-GAN: Semantic Image Commu- nications Using Masked Vector Quantization.IEEE Transactions on Cognitive Communications and Networking, 12:3363–3377
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Efficient Semantic Image Communication for Traffic Monitoring at the Edge
MMSD and SAMR achieve 99 percent and 99.1 percent average data reduction for traffic images by transmitting segmentation maps, edges, text or semantically masked JPEGs and reconstructing via diffusion or inpainting models.