Open Access

Fig. 7.

image

Download original image

Comparison of deep learning-based mass-mapping methods on one mock COSMOS field. First row: ground truth κTNG convergence map, the mean of our posterior samples (over 400 samples), DeepMass, and the standard deviation of our posterior samples. All the maps use the same color bar, except for the standard deviation, which is displayed in the range [0, 0.035]. Second row: samples from the posterior distribution, using the hybrid Gaussian-neural score prior. The input noise level is computed according Eq. (26) and the mask contours correspond to the COSMOS survey boundary mask.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.