Interactive Image Segmentation

Deep learning approach

click & scribble

• Deep Interactive Object Selection - N Xu et al., 2016 CVPR

  • positive and negative clicks
  • FCN + refinement by Graph Cut

• Regional Interactive Image Segmentation Networks - JH Liew et al., 2017 ICCV

  • click based
  • global branch + local regional refinement

• DeepIGeoS: A Deep Interactive Geodesic Framework for Medical Image Segmentation - G Wang et al., 2017 arXiv

  • click/ scribble based
  • Geodesic Distance Transformation
  • Refine Network

• Deep Extreme Cut: From Extreme Points to Object Segmentation - KK Maninis et al., 2018 CVPR

  • Four extreme points
  • [Pytorch code]

• Interactive Image Segmentation with Latent Diversity - Z L et al., 2018 CVPR

  • Minimizing the number of clicks
  • Multi-modality
  • Multiple choice learning

• Efficient Interactive Annotation of Segmentation Datasets with Polygon-RNN++ - D Acuna et al., 2018 CVPR

  • Polygon-RNN

natural language

• Guide Me: Interacting with Deep Networks - C Rupprecht et al., 2018 CVPR

weakly/semi supervised learning

• Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation - A Kolesnikov et al., 2016 ECCV

• What’s the Point: Semantic Segmentation with Point Supervision - A Bearman et al., 2016 ECCV

• Simple Does It: Weakly Supervised Instance and Semantic Segmentation - A Khoreva et al., 2017 CVPR

• Exploiting saliency for object segmentation from image level labels - SH Oh et al., 2017 CVPR

Graph based approach

• Interactive Graph Cuts for Optimal Boundary & Region Segmentation of Objects in N-D Images - YY Boykov et al., 2001 ICCV

• “GrabCut” — Interactive Foreground Extraction using Iterated Graph Cuts - C Rother et al., 2004 TOG

• Scribble-Supervised Convolutional Networks for Semantic Segmentation - D Lin et al., 2016 CVPR

Interactive Video Segmentation

Deep learning approach

natural language

• Video Object Segmentation with Language Referring Expressions - A Khoreva et al., 2018 arXiv