ASCL.net

Astrophysics Source Code Library

Making codes discoverable since 1999

ASCL Code Record

[submitted] Estimating photo-z of quasars based on a cross-modal contrastive learning method

MMLPhoto-z is a cross-modal contrastive learning approach for estimating photo-z of quasars. This method employs adversarial training and contrastive loss functions to promote the mutual conversion between multi-band photometric data features (magnitude, color) and photometric image features, while extracting modality-invariant features.

Code site:
https://github.com/Valar-hub/MMLPhoto-z

Views: 453