Inference of the optical depth to reionization τ from Planck CMB maps with convolutional neural networks

Open Access

Fig. 5.

Predictions of τ_NN from 10 000 simulations with input τ = 0.06, containing either CMB with Gaussian noise (left panels) or CMB with SRoll2 noise + systematics (right panels). The two rows denote different CNN models trained on CMB with Gaussian noise on a single frequency channel (top), on two frequency channels (bottom).

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.