EfficientSeg: An Efficient Semantic Segmentation Network
arXiv:2009.06469
Abstract
Deep neural network training without pre-trained weights and few data is shown to need more training iterations. It is also known that, deeper models are more successful than their shallow counterparts for semantic segmentation task. Thus, we introduce EfficientSeg architecture, a modified and scalable version of U-Net, which can be efficiently trained despite its depth. We evaluated EfficientSeg architecture on Minicity dataset and outperformed U-Net baseline score (40% mIoU) using the same parameter count (51.5% mIoU). Our most successful model obtained 58.1% mIoU score and got the fourth place in semantic segmentation track of ECCV 2020 VIPriors challenge.
References in corpus (11)
- Adam: A Method for Stochastic Optimization
- Rethinking Atrous Convolution for Semantic Image Segmentation
- Efficient Multi-Scale 3D CNN with Fully Connected CRF for Accurate Brain Lesion Segmentation
- GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classification
- Rethinking Pre-training and Self-training
- Some Improvements on Deep Convolutional Neural Network Based Image Classification
- Hierarchical Multi-Scale Attention for Semantic Segmentation
- Data Augmentation Using GANs
- Data Augmentation in Emotion Classification Using Generative Adversarial Networks
- Global Aggregation then Local Distribution in Fully Convolutional Networks
- HarDNet: A Low Memory Traffic Network