Advanced Deep Networks for 3D Mitochondria Instance Segmentation

16 Apr 2021  ·  Mingxing Li, Chang Chen, Xiaoyu Liu, Wei Huang, Yueyi Zhang, Zhiwei Xiong ·

Mitochondria instance segmentation from electron microscopy (EM) images has seen notable progress since the introduction of deep learning methods. In this paper, we propose two advanced deep networks, named Res-UNet-R and Res-UNet-H, for 3D mitochondria instance segmentation from Rat and Human samples. Specifically, we design a simple yet effective anisotropic convolution block and deploy a multi-scale training strategy, which together boost the segmentation performance. Moreover, we enhance the generalizability of the trained models on the test set by adding a denoising operation as pre-processing. In the Large-scale 3D Mitochondria Instance Segmentation Challenge at ISBI 2021, our method ranks the 1st place. Code is available at

PDF Abstract


Results from the Paper

Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
3D Instance Segmentation MitoEM Res-UNet-R/H AP75-R-Val 0.917 # 1
AP75-H-Val 0.828 # 1
AP75-R-Test 0.851 # 1
AP75-H-Test 0.829 # 1