Skip to content

MahmudulAlam/Recurrently-Semantic-Segmentation

Repository files navigation

Recurrently Semantic Segmentation

In this paper, a recurrent method of semantic segmentation is proposed where convolutional LSTM is used to generate a mask of objects present in the image recurrently. The proposed method only generates a mask of objects that are only present in the image. Therefore, requiring less output tensor space. The method uses Unet architecture and convolutional LSTM in the bottleneck and is trained using the PASCAL VOC dataset.

Requirements

Downloads

  • PyTorch==1.8.1
  • Weights: Download the pre-trained weights file of the recurrent semantic segmentation model and put the weights folder in the working directory.

Preview

Network Architecture

The proposed method is implemented using the following network architecture where encoder and decoder are used from U-net architecture and convolutional LSTM is used in the bottleneck.

Results

The experimental results are reported in terms of accuracy and mean interaction over union (IOU) between ground truth and predicted mask. The mean accuracy among all the classes is 30.02% and the mean IOU achieved among all the classes is 43.77%, and the maximum accuracy and IOU achieved are 70% and 71.54%, subsequently.

About

Recurrent method of semantic segmentation using convolutional LSTM

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published