The One Hundred Layers Tiramisu: Fully Convolutional DenseNets for Semantic Segmentation
arXiv:1611.09326
Abstract
State-of-the-art approaches for semantic image segmentation are built on Convolutional Neural Networks (CNNs). The typical segmentation architecture is composed of (a) a downsampling path responsible for extracting coarse semantic features, followed by (b) an upsampling path trained to recover the input image resolution at the output of the model and, optionally, (c) a post-processing module (e.g. Conditional Random Fields) to refine the model predictions. Recently, a new CNN architecture, Densely Connected Convolutional Networks (DenseNets), has shown excellent results on image classification tasks. The idea of DenseNets is based on the observation that if each layer is directly connected to every other layer in a feed-forward fashion then the network will be more accurate and easier to train. In this paper, we extend DenseNets to deal with the problem of semantic segmentation. We achieve state-of-the-art results on urban scene benchmark datasets such as CamVid and Gatech, without any further post-processing module nor pretraining. Moreover, due to smart construction of the model, our approach has much less parameters than currently published best entries for these datasets. Code to reproduce the experiments is available here : https://github.com/SimJeg/FC-DenseNet/blob/master/train.py
References in corpus (4)
Cited by in corpus (31)
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- Evolution of Image Segmentation using Deep Convolutional Neural Network: A Survey
- Weakly Supervised Medical Diagnosis and Localization from Multiple Resolutions
- DSOD: Learning Deeply Supervised Object Detectors from Scratch
- Stacked Deconvolutional Network for Semantic Segmentation
- Fully Convolutional Networks for Chip-wise Defect Detection Employing Photoluminescence Images
- Image to Image Translation for Domain Adaptation
- Models Matter, So Does Training: An Empirical Study of CNNs for Optical Flow Estimation
- DARTS: DenseUnet-based Automatic Rapid Tool for brain Segmentation
- High-Quality Face Image SR Using Conditional Generative Adversarial Networks
- Shape-from-Mask: A Deep Learning Based Human Body Shape Reconstruction from Binary Mask Images
- Deep learning for semantic segmentation of remote sensing images with rich spectral content
- Semantic Segmentation of Human Thigh Quadriceps Muscle in Magnetic Resonance Images
- A Unified Framework for Generalizable Style Transfer: Style and Content Separation
- Beyond Forward Shortcuts: Fully Convolutional Master-Slave Networks (MSNets) with Backward Skip Connections for Semantic Segmentation
- Model Generalization in Deep Learning Applications for Land Cover Mapping
- Joint Segmentation and Uncertainty Visualization of Retinal Layers in Optical Coherence Tomography Images using Bayesian Deep Learning
- Concatenated Feature Pyramid Network for Instance Segmentation
- Outline Objects using Deep Reinforcement Learning
- Stacked Neural Networks for end-to-end ciliary motion analysis
- Deep Geodesic Learning for Segmentation and Anatomical Landmarking
- Object Detection from Scratch with Deep Supervision
- Learning image from projection: a full-automatic reconstruction (FAR) net for sparse-views computed tomography
- Substitute Teacher Networks: Learning with Almost No Supervision
- Integrated Face Analytics Networks through Cross-Dataset Hybrid Training
- TLGAN: document Text Localization using Generative Adversarial Nets
- Semantic Segmentation via Highly Fused Convolutional Network with Multiple Soft Cost Functions
- ESFNet: Efficient Network for Building Extraction from High-Resolution Aerial Images
- Efficient Video Understanding via Layered Multi Frame-Rate Analysis
- Exploring Temporal Information for Improved Video Understanding
- Rethinking Fully Convolutional Networks for the Analysis of Photoluminescence Wafer Images