Abstract
Neural Architecture Search (NAS) has drawn significant attention as a tool for automatically constructing deep neural networks. The generated neural networks are mainly applied for image classification, and natural language processing. However, there are increasing demands for image segmentation in various areas, such as medical image processing, satellite image object location, and autopilot technology. We propose a NAS method called Automated Segmentation Network (AutoSegNet), targeting industrial and medical image segmentation. The search architectures are constructed by stacking the downsampling layer, the bridge layer, and the upsampling layer, which are explored by a recurrent neural network. Compared with other related methods for image segmentation, the proposed method has a small search space but can explore most of the-state-of-the-art supervised image segmentation models. We perform verification on two datasets, and the results show that AutoSegNet achieves superior segmentation results with clear and continuous segmented edges, as well as better image details.
Highlights
Neural Architecture Search (NAS) aims to search for the best neural network architecture, given the learning dataset
NAS-Unet [24] searches a U-like backbone network for medical image segmentation, and V-NAS [25] formulates the structure learning as differentiable neural architecture search, allowing the network to choose among 2D, 3D or Pseudo-3D (P3D) convolutions at each layer
The key idea of the proposed method is by searching the downsampling layer, the bridge layer, and the upsampling layer with an recurrent neural network(RNN) controller, and the best neural network architecture targeting on image segmentation given learning data can be discovered by the AutoSegNet
Summary
Neural Architecture Search (NAS) aims to search for the best neural network architecture, given the learning dataset. It has been successfully applied for image classification and language modeling [1]–[4]. Efficient Neural Architecture Search (ENAS) [3], targeting on image classification, generates the best neural network architecture with a fixed structure. The downsampling layer reduces the input size so that the network can learn from the more significant receptive field. Z. Xu et al.: AutoSegNet: An Automated Neural Network for Image Segmentation encoder-decoder network structure with the downsampling layers reducing the input size, the pooling operation is removed from the search space. The proposed method is tested on an industrial segmentation dataset as well as a medical segmentation dataset, and the results show quality segmentation with clear and continuous segmented edges and better details in the segmentation
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
