Vision18h ago
Tessera Launches Pixel-Wise Earth Observation Model
The Tessera foundation model provides pixel-level semantic segmentation for satellite imagery. It leverages a masked autoencoder architecture to learn spatial representations from diverse planetary data. This approach reduces the need for labeled datasets in environmental monitoring. Practitioners can now deploy more accurate land-cover classification tools with significantly less manual supervision.