← All models
FEVER Coming soon

FEVER - Fourier Encoder for Visual Embedding and Retrieval

Text and image retrieval

The first truly cross resolution retrieval model, a 341M FELA encoder that matches text to images and images to images across a wide resolution gap where perceptual hashing fails. Sold as a model and a bundled retrieval system, via self deploy and API.

License Lowdown Labs Lovely License 1.0
API accessComing soon
Docker / self-deployComing soon