# Stateless actor-critic for instance segmentation with high-level priors

arXiv: https://arxiv.org/abs/2107.02600 · HTML: https://mlsanity.com/paper/2107.02600/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 30 of its 56 references that are in the map.
- First posted on arXiv 6 Jul 2021.

## Builds on (most cited first, 30 of 30)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [Mask R-CNN](https://mlsanity.com/paper/1703.06870/) (2017)
- [Neural Message Passing for Quantum Chemistry](https://mlsanity.com/paper/1704.01212/) (2017)
- [Soft Actor-Critic Algorithms and Applications](https://mlsanity.com/paper/1812.05905/) (2018)
- [Value-Decomposition Networks For Cooperative Multi-Agent Learning](https://mlsanity.com/paper/1706.05296/) (2017)
- [Topology-Preserving Deep Image Segmentation](https://mlsanity.com/paper/1906.05404/) (2019)
- [Cell Detection with Star-convex Polygons](https://mlsanity.com/paper/1806.03535/) (2018)
- [Semantic Instance Segmentation with a Discriminative Loss Function](https://mlsanity.com/paper/1708.02551/) (2017)
- [Anatomically Constrained Neural Networks (ACNN): Application to Cardiac Image Enhancement and Segmentation](https://mlsanity.com/paper/1705.08302/) (2017)
- [Accelerated Hierarchical Density Clustering](https://mlsanity.com/paper/1705.07321/) (2017)
- [Instance Segmentation by Jointly Optimizing Spatial Embeddings and Clustering Bandwidth](https://mlsanity.com/paper/1906.11109/) (2019)
- [SSAP: Single-Shot Instance Segmentation With Affinity Pyramid](https://mlsanity.com/paper/1909.01616/) (2019)
- [Superhuman Accuracy on the SNEMI3D Connectomics Challenge](https://mlsanity.com/paper/1706.00120/) (2017)
- [Constrained-CNN losses for weakly supervised segmentation](https://mlsanity.com/paper/1805.04628/) (2018)
- [Connecting Generative Adversarial Networks and Actor-Critic Methods](https://mlsanity.com/paper/1610.01945/) (2016)
- [Scalable Learning of Non-Decomposable Objectives](https://mlsanity.com/paper/1608.04802/) (2016)
- [Scaling Wide Residual Networks for Panoptic Segmentation](https://mlsanity.com/paper/2011.11675/) (2020)
- [Training Deep Neural Networks via Direct Loss Minimization](https://mlsanity.com/paper/1511.06411/) (2015)
- [Continuous Dice Coefficient: a Method for Evaluating Probabilistic Segmentations](https://mlsanity.com/paper/1906.11031/) (2019)
- [Learning Surrogate Losses](https://mlsanity.com/paper/1905.10108/) (2019)
- [A Survey on Hough Transform, Theory, Techniques and Applications](https://mlsanity.com/paper/1502.02160/) (2015)
- [Sparse Object-level Supervision for Instance Segmentation with Pixel Embeddings](https://mlsanity.com/paper/2103.14572/) (2021)
- [Combinatorial Optimization for Panoptic Segmentation: A Fully Differentiable Approach](https://mlsanity.com/paper/2106.03188/) (2021)
- [End-to-end Learning for Graph Decomposition](https://mlsanity.com/paper/1812.09737/) (2018)
- [Actor-Critic Instance Segmentation](https://mlsanity.com/paper/1904.05126/) (2019)
- [Learning Loss Functions for Semi-supervised Learning via Discriminative Adversarial Networks](https://mlsanity.com/paper/1707.02198/) (2017)
- [Leveraging Domain Knowledge to Improve Microscopy Image Segmentation with Lifted Multicuts](https://mlsanity.com/paper/1905.10535/) (2019)
- [Semi-Automatic Generation of Tight Binary Masks and Non-Convex Isosurfaces for Quantitative Analysis of 3D Biological Samples](https://mlsanity.com/paper/2001.11469/) (2020)
- [Combinatorial Energy Learning for Image Segmentation](https://mlsanity.com/paper/1506.04304/) (2015)
- [Zero-shot Policy Learning with Spatial Temporal RewardDecomposition on Contingency-aware Observation](https://mlsanity.com/paper/1910.08143/) (2019)

## Abstract (arXiv)

Instance segmentation is an important computer vision problem which remains challenging despite impressive recent advances due to deep learning-based methods. Given sufficient training data, fully supervised methods can yield excellent performance, but annotation of ground-truth data remains a major bottleneck, especially for biomedical applications where it has to be performed by domain experts. The amount of labels required can be drastically reduced by using rules derived from prior knowledge to guide the segmentation. However, these rules are in general not differentiable and thus cannot be used with existing methods. Here, we relax this requirement by using stateless actor critic reinforcement learning, which enables non-differentiable rewards. We formulate the instance segmentation problem as graph partitioning and the actor critic predicts the edge weights driven by the rewards, which are based on the conformity of segmented instances to high-level priors on object shape, position or size. The experiments on toy and real datasets demonstrate that we can achieve excellent performance without any direct supervision based only on a rich set of priors.