Open Models Almanac
Methodsmodel architecturemultimodal architecture

single mention · model architecture

Multi-resolution visual feature fusion

Fuses low-level spatial information and higher-level semantic information from visual features at different resolutions.

The sources on this page only mention it in passing. The line above is the extraction's one-line gloss. See multimodal architecture for the methods filed beside it.

Also called multi-resolution visual features.

Filed under model architecture › multimodal architecture