CEREALS - Cost-Effective REgion-based Active Learning for Semantic Segmentation

Abstract

State of the art methods for semantic image segmentation are trained in a supervised fashion using a large corpus of fully labeled training images. However, gathering such a corpus is expensive, due to human annotation effort, in contrast to gathering unlabeled data. We propose an active learning-based strategy, called CEREALS, in which a human only has to hand-label a few, automatically selected, regions within an unlabeled image corpus. This minimizes human annotation effort while maximizing the performance of a semantic image segmentation method. The automatic selection procedure is achieved by: a) using a suitable information measure combined with an estimate about human annotation effort, which is inferred from a learned cost model, and b) exploiting the spatial coherency of an image. The performance of CEREALS is demonstrated on Cityscapes, where we are able to reduce the annotation effort to 17%, while keeping 95% of the mean Intersection over Union (mIoU) of a model that was trained with the fully annotated training set of Cityscapes.

arXiv commentPublished at British Machine Vision Conference 2018 (BMVC)

References in corpus (15)

• MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
• The Cityscapes Dataset for Semantic Urban Scene Understanding
• Deep Bayesian Active Learning with Image Data
• Adversarial Learning for Semi-Supervised Semantic Segmentation
• What makes ImageNet good for transfer learning?
• Revisiting Unreasonable Effectiveness of Data in Deep Learning Era
• Playing for Data: Ground Truth from Computer Games
• Can Image-Level Labels Replace Pixel-Level Labels for Image Parsing
• Deep Interactive Object Selection
• Annotating Object Instances with a Polygon-RNN
• Suggestive Annotation: A Deep Active Learning Framework for Biomedical Image Segmentation
• Exploiting saliency for object segmentation from image level labels
• Weakly Supervised Semantic Segmentation using Web-Crawled Videos
• Built-in Foreground/Background Prior for Weakly-Supervised Semantic Segmentation
• Semantic Instance Annotation of Street Scenes by 3D to 2D Label Transfer

Cited by in corpus (6)

• Structured Consistency Loss for semi-supervised semantic segmentation
• Reinforced active learning for image segmentation
• A Weakly Supervised Region-Based Active Learning Method for COVID-19 Segmentation in CT Images
• Deep Active Learning for Efficient Training of a LiDAR 3D Object Detector
• Efficacy of Pixel-Level OOD Detection for Semantic Segmentation
• Are We Hungry for 3D LiDAR Data for Semantic Segmentation? A Survey and Experimental Study