A high-speed multi-GPU implementation of bottom-up attention using CUDA.

Tingting Xu, Thomas Pototschnig, Kolja Khnlenz, Martin Buss

Browse the full ICRA paper archive.