An efficient solution for semantic segmentation: ShuffleNet V2 with atrous separable convolutions
arXiv:1902.07476 · doi:10.1007/978-3-030-20205-7_4
Abstract
Assigning a label to each pixel in an image, namely semantic segmentation, has been an important task in computer vision, and has applications in autonomous driving, robotic navigation, localization, and scene understanding. Fully convolutional neural networks have proved to be a successful solution for the task over the years but most of the work being done focuses primarily on accuracy. In this paper, we present a computationally efficient approach to semantic segmentation, while achieving a high mean intersection over union (mIOU), 70.33% on Cityscapes challenge. The network proposed is capable of running real-time on mobile devices. In addition, we make our code and model weights publicly available.
12 pages, 6 figures, 5 tables
References in corpus (8)
- Adam: A Method for Stochastic Optimization
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- Rethinking Atrous Convolution for Semantic Image Segmentation
- Fully Convolutional Networks for Semantic Segmentation
- ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation
- Searching for Efficient Multi-Scale Architectures for Dense Image Prediction
- ShuffleSeg: Real-time Semantic Segmentation Network