| DC Field | Value | Language |
|---|
| dc.contributor.author | Adaska, E. | - |
| dc.contributor.author | Lechanka, A. | - |
| dc.date.accessioned | 2021-11-04T05:51:01Z | - |
| dc.date.available | 2021-11-04T05:51:01Z | - |
| dc.date.issued | 2021 | - |
| dc.identifier.citation | Adaska, E. UNetX: Real-time Pedestrian Crosswalk Segmentation on Mobile Device / Adaska E., Lechanka A. // Pattern Recognition and Information Processing (PRIP'2021) = Распознавание образов и обработка информации (2021) : Proceedings of the 15th International Conference, 21–24 Sept. 2021, Minsk, Belarus / United Institute of Informatics Problems of the National Academy of Sciences of Belarus. – Minsk, 2021. – P. 75–78. | ru_RU |
| dc.identifier.uri | https://libeldoc.bsuir.by/handle/123456789/45774 | - |
| dc.description.abstract | This paper presents a lightweight deep neural network that segments pedestrian crosswalks on an image in realtime. It is based on U-Net architecture with all its convolution layers substituted with depthwise separable convolution ones. This neural network was trained and tested against a set of manually segmented 3083 road images — with and without crosswalks. The resulting network has only 383K parameters and runs at 35 FPS on a mobile phone. The Jaccard index (IoUmetric) on the validation set is 0.9138. | ru_RU |
| dc.language.iso | en | ru_RU |
| dc.publisher | UIIP NASB | ru_RU |
| dc.subject | материалы конференций | ru_RU |
| dc.subject | conference proceedings | ru_RU |
| dc.subject | autonomous car | ru_RU |
| dc.subject | pedestrian crosswalk | ru_RU |
| dc.subject | image segmentation | ru_RU |
| dc.subject | deep neural network | ru_RU |
| dc.subject | U-Net | ru_RU |
| dc.subject | depthwise separable convolution | ru_RU |
| dc.title | UNetX: Real-time Pedestrian Crosswalk Segmentation on Mobile Device | ru_RU |
| dc.type | Статья | ru_RU |
| Appears in Collections: | Pattern Recognition and Information Processing (PRIP'2021) = Распознавание образов и обработка информации (2021)
|
