Batch Size Influence on Performance of Graphic and Tensor Processing Units during Training and Inference Phases

Autor: Kochura, Yuriy, Gordienko, Yuri, Taran, Vlad, Gordienko, Nikita, Rokovyi, Alexandr, Alienin, Oleg, Stirenko, Sergii

Publikováno v: Hu Z., Petoukhov S., Dychka I., He M. (eds) Advances in Computer Science for Engineering and Education II. ICCSEEA 2019. Advances in Intelligent Systems and Computing, vol 938 (pp. 658-668). Springer, Cham

The impact of the maximally possible batch size (for the better runtime) on performance of graphic processing units (GPU) and tensor processing units (TPU) during training and inference phases is investigated. The numerous runs of the selected deep n

Externí odkaz: http://arxiv.org/abs/1812.11731

Performance Evaluation of Deep Learning Networks for Semantic Segmentation of Traffic Stereo-Pair Images

Autor: Taran, Vlad, Gordienko, Nikita, Kochura, Yuriy, Gordienko, Yuri, Rokovyi, Alexandr, Alienin, Oleg, Stirenko, Sergii

Publikováno v: Proceedings of the 19th International Conference on Computer Systems and Technologies (CompSysTech'18), Boris Rachev and Angel Smrikarov (Eds.). ACM, New York, NY, USA, 73-80 (2018)

Semantic image segmentation is one the most demanding task, especially for analysis of traffic conditions for self-driving cars. Here the results of application of several deep learning architectures (PSPNet and ICNet) for semantic image segmentation

Externí odkaz: http://arxiv.org/abs/1806.01896