On uncertainty estimation in active learning for image segmentation

Uncertainty estimation is important for interpreting the trustworthiness of machine learning models in many applications. This is especially critical in the data-driven active learning setting where the goal is to achieve a certain accuracy with minimum labeling effort. In such settings, the model learns to select the most informative unlabeled samples for annotation based on its estimated uncertainty. The highly uncertain predictions are assumed to be more informative for improving model performance. In this paper, we explore uncertainty calibration within an active learning framework for medical image segmentation, an area where labels often are scarce. Various uncertainty estimation methods and acquisition strategies (regions and full images) are investigated. We observe that selecting regions to annotate instead of full images leads to more well-calibrated models. Additionally, we experimentally show that annotating regions can cut 50% of pixels that need to be labeled by humans compared to annotating full images.

[1]  Bai Ying Lei,et al.  Accurate Segmentation of Cervical Cytoplasm and Nuclei Based on Multiscale Convolutional Network and Graph Partitioning , 2015, IEEE Transactions on Biomedical Engineering.

[2]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[3]  Zhuowen Tu,et al.  Training Deeper Convolutional Networks with Deep Supervision , 2015, ArXiv.

[4]  Zhuowen Tu,et al.  Deeply-Supervised Nets , 2014, AISTATS.

[5]  Noel C. F. Codella,et al.  Skin lesion analysis toward melanoma detection: A challenge at the 2017 International symposium on biomedical imaging (ISBI), hosted by the international skin imaging collaboration (ISIC) , 2016, 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018).

[6]  Carl E. Rasmussen,et al.  Evaluating Predictive Uncertainty Challenge , 2005, MLCW.

[7]  Yang Li,et al.  Gland Instance Segmentation by Deep Multichannel Side Supervision , 2016, MICCAI.

[8]  Jian Sun,et al.  Identity Mappings in Deep Residual Networks , 2016, ECCV.

[9]  Julien Valentin,et al.  ViewAL: Active Learning With Viewpoint Entropy for Semantic Segmentation , 2020, 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Andreas Nürnberger,et al.  The Power of Ensembles for Active Learning in Image Classification , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[11]  Yang Li,et al.  Gland Instance Segmentation Using Deep Multichannel Neural Networks , 2016, IEEE Transactions on Biomedical Engineering.

[12]  Hossam Faris,et al.  Empirical Evaluation of the Cycle Reservoir with Regular Jumps for Time Series Forecasting: A Comparison Study , 2017, CSOC.

[13]  Zoubin Ghahramani,et al.  Bayesian Active Learning for Classification and Preference Learning , 2011, ArXiv.

[14]  Sebastian Nowozin,et al.  Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift , 2019, NeurIPS.

[15]  Pedro H. O. Pinheiro,et al.  Reinforced active learning for image segmentation , 2020, ICLR.

[16]  Hao Chen,et al.  Gland segmentation in colon histology images: The glas challenge contest , 2016, Medical Image Anal..

[17]  Zoubin Ghahramani,et al.  Deep Bayesian Active Learning with Image Data , 2017, ICML.

[18]  David A. Cohn,et al.  Active Learning with Statistical Models , 1996, NIPS.

[19]  Hao Chen,et al.  DCAN: Deep Contour-Aware Networks for Accurate Gland Segmentation , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[20]  Gopinath Chennupati,et al.  On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networks , 2019, NeurIPS.

[21]  Zoubin Ghahramani,et al.  Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning , 2015, ICML.

[22]  Ling Zhang,et al.  A deep learning based framework for accurate segmentation of cervical cytoplasm and nuclei , 2014, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[23]  Ghassan Hamarneh,et al.  Multi-loss convolutional networks for gland analysis in microscopy , 2016, 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI).

[24]  Kilian Q. Weinberger,et al.  On Calibration of Modern Neural Networks , 2017, ICML.

[25]  Roman Jakubicek,et al.  Deep convolutional neural network‐based segmentation and classification of difficult to define metastatic spinal lesions in 3D CT data , 2018, Medical Image Anal..

[26]  Mitko Veta,et al.  Cutting Out the Middleman: Measuring Nuclear Area in Histopathology Slides Without Segmentation , 2016, MICCAI.

[27]  Lin Yang,et al.  Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentation , 2017, MICCAI.

[28]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[29]  Carsten Rother,et al.  CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentation , 2018, BMVC.