Name
Abstract | ||
|---|---|---|
Recently, convolutional neural networks (CNNs) have made a big splash in the field of semantic segmentation, achieving very high segmentation accuracy. In order to meet the requirement of real-time inference, existing methods increase inference speed by reducing the image resolution, leading to lower segmentation performance. We propose in this work a multi-level feature fusion network referred to as MLFFNet that utilizes a novel deep neural network architecture for efficient and real-time semantic segmentation. To strike a balance between speed and performance, MLFFNet substantially reduces the computational complexity by using a lightweight feature extraction network to implement feature reuse through multi-level feature fusion. In addition, MLFFNet targets at excellent segmentation performance through a channel attention mechanism and dilated convolutions with different rates. Specifically, MLFFNet achieves 72.6% mIoU on Cityscapes with the speed of 68.3 FPS on one NVIDIA Titan X card, which is significantly faster than the existing methods with comparable performance. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/WCSP.2019.8927880 | 2019 11th International Conference on Wireless Communications and Signal Processing (WCSP) |
Keywords | Field | DocType |
real-time semantic segmentation,feature fusion,channel attention,dilated convolution | Pattern recognition,Convolutional neural network,Computer science,Inference,Segmentation,Communication channel,Real-time computing,Feature extraction,Image segmentation,Artificial intelligence,Decoding methods,Computational complexity theory | Conference |
ISSN | ISBN | Citations |
2325-3746 | 978-1-7281-3556-4 | 0 |
PageRank | References | Authors |
0.34 | 3 | 5 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Lu Wang | 1 | 0 | 0.34 |
| Qinzhen Xu | 2 | 0 | 0.34 |
| Zixiang Xiong | 3 | 0 | 0.34 |
| Yongming Huang | 4 | 1472 | 146.50 |
| Luxi Yang | 5 | 1180 | 118.08 |