atrous

From āter (“dark, black”) +‎ -ous.

• IPA^(key): /ˈeɪtɹəs/

atrous (comparative more atrous, superlative most atrous)

1. jet-black in color.
   • 1830, Jonathan Stokes, Botanical Commentaries- Volume 1:

     Male atrous leadcolourd^[sic] on the upper side, under side a mixture of whitish and ferruginous with cuspidate fasciae ;

   • 1917, Proceedings of the Linnean Society of New South Wales:

     Head atrous, viridescent on lateral frontal spaces and occiput, gulae, basal part of mentum on each side of median tooth brightly viridescent; prothorax atrous, with viridescent tints on disc and wide brassy-green basal and lateral margins.

   • 1968, Iconographia Mycologica - Issues 22-27:

     Perithecia sparse, innate, then erumpent for apical portion, subglobose, atrous, with ostiollar portion a little enlarged.

   • 2013, DB Angel, The Gift, page 13:

     An imposing portrait of a Caucasian male holding an atrous book with a sheepish smile hung ominously over the bed.

   • 2014, Mark Alan Lindsley, Future Earth, page 405:

     The throng shouted, female Eligor flung flowers from the roofs as the rhythmic chiming of iron-shod riding beasts came clearer, and the first of the mighty array swung into view in the broad, atrous streets that curved round the spired Grand Union of Imperial Opal.

From French à trous (“with holes”).

• IPA^(key): /əˈtruː/

atrous (not comparable)

1. (computing, rare) dilated (used when describing convolutions)

   Atrous convolution allows us to explicitly control the resolution.

   • 2008, Sharon Priya and A. Srinivasan, “Image and Video Denoising by Sparse 3-D Transform Domain Using Atrous and Gabor Wavelets”, in Proceedings Second International Conference on Information Processing, page 87:

     The new approach is executed by using atrous and gabor wavelets that work effectively on image and video.

   • 2019, Qi Wang, Learning to Understand Remote Sensing Images, page 158:

     In order to further illustrate the effect of "atrous" convolution, we compare it with standard convolution using a simple example in Figure 4.

   • 2019, Guanhong Zhang, Muyi Sun, and Xiaoguang Zhou, “Sub-pixel Upsampling Decode Network for Semantic Segmentation”, in Artificial Intelligence, page 18:

     As the frameworks based on atrous convolution and ASPP module have proved their success, we implement an encoder with similar structure of [5], using a pretrained ResNetV2 as backbone network.

• Austro-, Satoru, Soutar, soutar, stauro-