TEGLIE: Transformer encoders as strong gravitational lens finders in KiDS - From simulations to surveys

Open Access

Table 3

Test and training set sizes for the different kinds of data augmentations.

Augmentation type	Test set	Training set
No Augmentation	68	243
Rotated (R)	271	972
Rotated+ $N$ $\[\boldsymbol{\mathcal{N}}\]$ (0, 10⁻¹¹)	338	1216
Rotated+ $N$ $\[\boldsymbol{\mathcal{N}}\]$ (0, 10⁻¹⁵)	338	1216
Rotated, Rotated+ $N$ $\[\boldsymbol{\mathcal{N}}\]$ (0, 10⁻¹¹)	541	1946
Rotated, Flipped (RF)	340	1222
Rotated, Flipped, Transposed (RFT)	406	1459

Notes. The dataset size includes lenses and not lenses together.

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.