Label-PEnet: Sequential Label Propagation and Enhancement Networks for Weakly Supervised Instance Segmentation

Weifeng Ge, Sheng Guo, Weilin Huang, Matthew R. Scott; The IEEE International Conference on Computer Vision (ICCV), 2019, pp. 3345-3354

Abstract


Weakly-supervised instance segmentation aims to detect and segment object instances precisely, given image-level labels only. Unlike previous methods which are composed of multiple offline stages, we propose Sequential Label Propagation and Enhancement Networks (referred as Label-PEnet) that progressively transforms image-level labels to pixel-wise labels in a coarse-to-fine manner. We design four cascaded modules including multi-label classification, object detection, instance refinement and instance segmentation, which are implemented sequentially by sharing the same backbone. The cascaded pipeline is trained alternatively with a curriculum learning strategy that generalizes labels from high level images to low-level pixels gradually with increasing accuracy. In addition, we design a proposal calibration module to explore the ability of classification networks to find key pixels that identify object parts, which serves as a post validation strategy running in the inverse order. We evaluate the efficiency of our Label-PEnet in mining instance masks on standard benchmarks: PASCAL VOC 2007 and 2012. Experimental results show that Label-PEnet outperforms the state-of-art algorithms by a clear margin, and obtains comparable performance even with fully supervised approaches.

Related Material


[pdf]
[bibtex]
@InProceedings{Ge_2019_ICCV,
author = {Ge, Weifeng and Guo, Sheng and Huang, Weilin and Scott, Matthew R.},
title = {Label-PEnet: Sequential Label Propagation and Enhancement Networks for Weakly Supervised Instance Segmentation},
booktitle = {The IEEE International Conference on Computer Vision (ICCV)},
month = {October},
year = {2019}
}