An End-to-End Tree Based Approach for Instance Segmentation

Abstract

This paper presents an approach for bottom-up hierarchical instance segmentation. We propose an end-to-end model to estimate energies of regions in an hierarchical region tree. To this end, we introduce a Convolutional Tree-LSTM module to leverage the tree-structured network topology. For constructing the hierarchical region tree, we utilize the accurate boundaries predicted from a pre-trained convolutional oriented boundary network. We evaluate our model on PASCAL VOC 2012 dataset showing that we obtain good trade-off between segmentation accuracy and time taken to process a single image.

Cite

Text

Manohar and Niitani. "An End-to-End Tree Based Approach for Instance Segmentation." European Conference on Computer Vision Workshops, 2018. doi:10.1007/978-3-030-11021-5_30

Markdown

[Manohar and Niitani. "An End-to-End Tree Based Approach for Instance Segmentation." European Conference on Computer Vision Workshops, 2018.](https://mlanthology.org/eccvw/2018/manohar2018eccvw-endtoend/) doi:10.1007/978-3-030-11021-5_30

BibTeX

@inproceedings{manohar2018eccvw-endtoend,
  title     = {{An End-to-End Tree Based Approach for Instance Segmentation}},
  author    = {Manohar, K. V. and Niitani, Yusuke},
  booktitle = {European Conference on Computer Vision Workshops},
  year      = {2018},
  pages     = {521-527},
  doi       = {10.1007/978-3-030-11021-5_30},
  url       = {https://mlanthology.org/eccvw/2018/manohar2018eccvw-endtoend/}
}