# Consistent Recurrent Neural Networks for 3D Neuron Segmentation

arXiv: https://arxiv.org/abs/2102.01021 · HTML: https://mlsanity.com/paper/2102.01021/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 16 references that are in the map.
- First posted on arXiv 1 Feb 2021.

## Builds on (most cited first, 9 of 9)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting](https://mlsanity.com/paper/1506.04214/) (2015)
- [RVOS: End-to-End Recurrent Network for Video Object Segmentation](https://mlsanity.com/paper/1903.05612/) (2019)
- [Superhuman Accuracy on the SNEMI3D Connectomics Challenge](https://mlsanity.com/paper/1706.00120/) (2017)
- [Deep Bidirectional and Unidirectional LSTM Recurrent Neural Network for Network-wide Traffic Speed Prediction](https://mlsanity.com/paper/1801.02143/) (2018)
- [Image Segmentation by Size-Dependent Single Linkage Clustering of a Watershed Basin Graph](https://mlsanity.com/paper/1505.00249/) (2015)
- [Large Scale Image Segmentation with Structured Loss based Deep Learning for Connectome Reconstruction](https://mlsanity.com/paper/1709.02974/) (2017)
- [Flood-Filling Networks](https://mlsanity.com/paper/1611.00421/) (2016)
- [A Multi-Pass Approach to Large-Scale Connectomics](https://mlsanity.com/paper/1612.02120/) (2016)

## Built on by (most cited first, 3 of 3)

- [AxonEM Dataset: 3D Axon Instance Segmentation of Brain Cortical Regions](https://mlsanity.com/paper/2107.05451/) (2021)
- [VICE: Visual Identification and Correction of Neural Circuit Errors](https://mlsanity.com/paper/2105.06861/) (2021)
- [Mitigating Pretraining-Induced Attention Asymmetry in 2D+ Electron Microscopy Image Segmentation](https://mlsanity.com/paper/2505.14105/) (2025)

## Abstract (arXiv)

We present a recurrent network for the 3D reconstruction of neurons that sequentially generates binary masks for every object in an image with spatio-temporal consistency. Our network models consistency in two parts: (i) local, which allows exploring non-occluding and temporally-adjacent object relationships with bi-directional recurrence. (ii) non-local, which allows exploring long-range object relationships in the temporal domain with skip connections. Our proposed network is end-to-end trainable from an input image to a sequence of object masks, and, compared to methods relying on object boundaries, its output does not require post-processing. We evaluate our method on three benchmarks for neuron segmentation and achieved state-of-the-art performance on the SNEMI3D challenge.