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 with 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 the probabilities for each galaxy of having a spheroid, a disk, presenting an irregularity, being compact or point source and being unclassifiable. ConvNets are able to predict the fractions of votes given a galaxy image with zero bias and ∼10% scatter. The fraction of miss-classifications is less than 1%. Our classification scheme represents a major improvement with respect to CAS (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 \href
@article{arxiv.1509.05429,
title = {A catalog of visual-like morphologies in the 5 CANDELS fields using deep-learning},
author = {M. Huertas-Company and R. Gravet and G. Cabrera-Vives and P. G. Pérez-González and J. S. Kartaltepe and G. Barro and M. Bernardi and S. Mei and F. Shankar and P. Dimauro and E. F. Bell and D. Kocevski and D. C. Koo and S. M. Faber and D. H. Mcintosh},
journal= {arXiv preprint arXiv:1509.05429},
year = {2015}
}
Comments
Accepted for publication in ApjS. Figure 10 summarizes the excellent agreement between our classification and a pure visual one. Table 3 shows the content of the catalogs. The catalogs are available from the Rainbow database (http://rainbowx.fis.ucm.es/Rainbow_navigator_public) based on the selections from the CANDELS team and cross-matched with 3D-HST v4.1 catalogs