Bounding boxes for weakly supervised segmentation: Global constraints get close to full supervision

We propose a novel weakly supervised learning segmentation based on several global constraints derived from box annotations. Particularly, we leverage a classical tightness prior to a deep learning setting via imposing a set of constraints on the network outputs. Such a powerful topological prior prevents solutions from excessive shrinking by enforcing any horizontal or vertical line within the bounding box to contain, at least, one pixel of the foreground region. Furthermore, we integrate our deep tightness prior with a global background emptiness constraint, guiding training with information outside the bounding box. We demonstrate experimentally that such a global constraint is much more powerful than standard cross-entropy for the background class. Our optimization problem is challenging as it takes the form of a large set of inequality constraints on the outputs of deep networks. We solve it with sequence of unconstrained losses based on a recent powerful extension of the log-barrier method, which is well-known in the context of interior-point methods. This accommodates standard stochastic gradient descent (SGD) for training deep networks, while avoiding computationally expensive and unstable Lagrangian dual steps and projections. Extensive experiments over two different public data sets and applications (prostate and brain lesions) demonstrate that the synergy between our global tightness and emptiness priors yield very competitive performances, approaching full supervision and outperforming significantly DeepCut. Furthermore, our approach removes the need for computationally expensive proposal generation. Our code is shared anonymously.

[1]  Fei-Fei Li,et al.  What's the Point: Semantic Segmentation with Point Supervision , 2015, ECCV.

[2]  Eugenio Culurciello,et al.  ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation , 2016, ArXiv.

[3]  Zhipeng Jia,et al.  Constrained Deep Weak Supervision for Histopathology Image Segmentation , 2017, IEEE Transactions on Medical Imaging.

[4]  Bram van Ginneken,et al.  A survey on deep learning in medical image analysis , 2017, Medical Image Anal..

[5]  Jose Dolz,et al.  3D fully convolutional networks for subcortical segmentation in MRI: A large-scale study , 2016, NeuroImage.

[6]  Eric Granger,et al.  Curriculum semi-supervised segmentation , 2019, MICCAI.

[7]  Andrew Blake,et al.  "GrabCut" , 2004, ACM Trans. Graph..

[8]  Meritxell Bach Cuadra,et al.  A novel segmentation framework for uveal melanoma based on magnetic resonance imaging and class activation maps , 2018 .

[9]  Jian Sun,et al.  ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[11]  Vladlen Koltun,et al.  Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials , 2011, NIPS.

[12]  Jose Dolz,et al.  Constrained domain adaptation for segmentation , 2019, MICCAI.

[13]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[14]  Jian Sun,et al.  BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[15]  Dimitris N. Metaxas,et al.  Weakly Supervised Deep Nuclei Segmentation using Points Annotation in Histopathology Images , 2019, MIDL.

[16]  Bernt Schiele,et al.  Simple Does It: Weakly Supervised Instance and Semantic Segmentation , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Toby Sharp,et al.  Image segmentation with a bounding box prior , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[18]  Yan Huang,et al.  Box-Driven Class-Wise Region Masking and Filling Rate Guided Loss for Weakly Supervised Semantic Segmentation , 2019, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  Ismail Ben Ayed,et al.  On Regularized Losses for Weakly-supervised CNN Segmentation , 2018, ECCV.

[20]  George Papandreou,et al.  Weakly-and Semi-Supervised Learning of a Deep Convolutional Network for Semantic Image Segmentation , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[21]  Florian Jung,et al.  Evaluation of prostate segmentation algorithms for MRI: The PROMISE12 challenge , 2014, Medical Image Anal..

[22]  Trevor Darrell,et al.  Constrained Convolutional Neural Networks for Weakly Supervised Segmentation , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[23]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[24]  Jonathan T. Barron,et al.  Multiscale Combinatorial Grouping for Image Segmentation and Object Proposal Generation , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[25]  Jianfeng Feng,et al.  Weakly Supervised Brain Lesion Segmentation via Attentional Representation Learning , 2019, MICCAI.

[26]  Eric Granger,et al.  Constrained‐CNN losses for weakly supervised segmentation☆ , 2018, Medical Image Anal..

[27]  Qi Zou,et al.  GraphNet: Learning Image Pseudo Annotations for Weakly-Supervised Semantic Segmentation , 2018, ACM Multimedia.

[28]  Yung-Yu Chuang,et al.  Weakly Supervised Instance Segmentation using the Bounding Box Tightness Prior , 2019, NeurIPS.

[29]  Lei Ai,et al.  A large, open source dataset of stroke anatomical brain images and manual lesion segmentations , 2017, Scientific Data.

[30]  Ismail Ben Ayed,et al.  Beyond Gradient Descent for Regularized Segmentation Losses , 2018, 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).