Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding

Code

• github.com/hoidn/ptychopinn
  tf★ 22
• github.com/SkyWa7ch3r/ImageSegmentation
  tf★ 7
• github.com/yubaoliu/caffe-segnet
  none★ 0
• github.com/Zhanghongbin-github/SegNet-Tutorial
  caffe2★ 0
• github.com/hosshonarvar/Image-Segmentation
  tf★ 0
• github.com/alexgkendall/SegNet-Tutorial
  caffe2★ 0
• github.com/Fangrn/caffe-segnet
  none★ 0
• github.com/azy64/Deep-Learning
  tf★ 0
• github.com/SkyWa7ch3r/SceneSegmentation
  tf★ 0
• github.com/Paultool/segnet
  none★ 0

Abstract

We present a deep learning framework for probabilistic pixel-wise semantic segmentation, which we term Bayesian SegNet. Semantic segmentation is an important tool for visual scene understanding and a meaningful measure of uncertainty is essential for decision making. Our contribution is a practical system which is able to predict pixel-wise class labels with a measure of model uncertainty. We achieve this by Monte Carlo sampling with dropout at test time to generate a posterior distribution of pixel class labels. In addition, we show that modelling uncertainty improves segmentation performance by 2-3% across a number of state of the art architectures such as SegNet, FCN and Dilation Network, with no additional parametrisation. We also observe a significant improvement in performance for smaller datasets where modelling uncertainty is more effective. We benchmark Bayesian SegNet on the indoor SUN Scene Understanding and outdoor CamVid driving scenes datasets.

Tasks

Reproductions