We present a catalog of visual-like H-band morphologies of ~50.000 galaxies (Hf160w < 24.5) in the 5 CANDELS fields (GOODS-N, GOODS-S, UDS, EGS, and COSMOS). Morphologies are estimated using Convolutional Neural Networks (ConvNets). The median redshift of the sample is 〈z〉 ~ 1.25. The algorithm is trained on GOODS-S, for which visual classifications are publicly available, and then applied to the other 4 fields. Following the CANDELS main morphology classification scheme, our model retrieves for each galaxy the probabilities of having a spheroid or a disk, presenting an irregularity, being compact or a point source, and being unclassifiable. ConvNets are able to predict the fractions of votes given to a galaxy image with zero bias and ~10% scatter. The fraction of mis-classifications is less than 1%. Our classification scheme represents a major improvement with respect to Concentration-Asymmetry-Smoothness-based methods, which hit a 20%–30% contamination limit at high z. The catalog is released with the present paper via the Rainbow database (http://rainbowx.fis.ucm.es/Rainbow_navigator_public/).

Document Type


Publication Date


Notes/Citation Information

Published in The Astrophysical Journal Supplement Series, v. 221, no. 1, article 8, p. 1-23.

© 2015. The American Astronomical Society. All rights reserved.

The copyright holders have granted the permission for posting the article here.

Digital Object Identifier (DOI)


Funding Information

G.C.V gratefully acknowledges financial support from CONICYT-Chile through its doctoral scholarship and grant DPI20140090. S.M. acknowledges financial support from the Institut Universitaire de France (IUF), of which she is senior member. G.B., D.C.K., and S.M.F. acknowledge support from NSF grant AST-08-08133 and NASA grant HST-GO- 12060.10A.

Related Content

The catalog of ~50,000 galaxies is released with the present paper through the Rainbow database: http://rainbowx.fis.ucm.es/Rainbow_navigator_public/.