Iteratively Trained Interactive Segmentation

Abstract

Deep learning requires large amounts of training data to be effective. For the task of object segmentation, manually labeling data is very expensive, and hence interactive methods are needed. Following recent approaches, we develop an interactive object segmentation system which uses user input in the form of clicks as the input to a convolutional network. While previous methods use heuristic click sampling strategies to emulate user clicks during training, we propose a new iterative training strategy. During training, we iteratively add clicks based on the errors of the currently predicted segmentation. We show that our iterative training strategy together with additional improvements to the network architecture results in improved results over the state-of-the-art.

References in corpus (5)

• Rethinking Atrous Convolution for Semantic Image Segmentation
• The Cityscapes Dataset for Semantic Urban Scene Understanding
• Bridging Category-level and Instance-level Semantic Image Segmentation
• Deep Interactive Object Selection
• Annotating Object Instances with a Polygon-RNN

Cited by in corpus (6)

• Segment Anything Model for Medical Image Analysis: an Experimental Study
• EdgeFlow: Achieving Practical Interactive Segmentation with Edge-Guided Flow
• DISIR: Deep Image Segmentation with Interactive Refinement
• ScribbleBox: Interactive Annotation Framework for Video Object Segmentation
• Iterative Interaction Training for Segmentation Editing Networks
• A Fully Convolutional Two-Stream Fusion Network for Interactive Image Segmentation