| Title: | Low-latency Human Portrait Segmentation Network Optimized for CPU Inference |
| Authors: | Pirshtuk, D. |
| Keywords: | материалы конференций;portrait segmentation;image segmentation;neural networks;computer vision;augmented reality |
| Issue Date: | 2023 |
| Publisher: | BSU |
| Citation: | Pirshtuk, D. Low-latency Human Portrait Segmentation Network Optimized for CPU Inference / D. Pirshtuk // Pattern Recognition and Information Processing (PRIP'2023) = Распознавание образов и обработка информации (2023) : Proceedings of the 16th International Conference, October 17–19, 2023, Minsk, Belarus / United Institute of Informatics Problems of the National Academy of Sciences of Belarus. – Minsk, 2023. – P. 186–192. |
| Abstract: | In this paper, we discuss a design of fast and lightweight neural networks for working in real-time under
very strict resource constraints and describe a human portrait segmentation method with temporal consistency based on an encoder-decoder architecture with a state-of-the-art CPU optimized PP-LCNet backbone and a custom decoder. Proposed neural network can process about 150-500 frames per second using only a single CPU thread with high accuracy and can be used for virtual background replacements in video conferencing and other augmented reality cases. |
| URI: | https://libeldoc.bsuir.by/handle/123456789/54315 |
| Appears in Collections: | Pattern Recognition and Information Processing (PRIP'2023) = Распознавание образов и обработка информации (2023)
|
