Datasets:
Search is not available for this dataset
main array 2D |
|---|
[
[
128,
137,
141,
147,
143,
145,
140,
144,
140,
138,
143,
132,
138,
136,
134,
133,
134,
131,
132,
137,
142,
135,
139,
146,
139,
148,
147,
139,
149,
155,
143,
141,
140,
143,
147,... |
[[147,146,145,144,144,150,149,152,148,142,141,144,137,132,133,134,130,134,139,140,135,135,131,139,13(...TRUNCATED) |
[[160,154,150,154,149,151,149,148,148,153,146,144,135,140,129,134,140,132,129,136,137,137,134,137,13(...TRUNCATED) |
[[148,145,151,153,151,153,144,153,148,150,143,142,134,137,137,149,139,124,145,149,139,134,143,144,13(...TRUNCATED) |
[[156,156,153,150,151,159,152,155,153,153,144,134,136,141,131,129,141,135,133,146,135,137,135,138,13(...TRUNCATED) |
[[156,157,156,155,156,144,158,149,154,153,143,139,145,139,135,140,144,138,140,152,149,139,133,131,13(...TRUNCATED) |
[[161,167,160,163,156,150,153,152,156,154,147,139,146,141,139,139,145,147,141,147,151,139,140,139,13(...TRUNCATED) |
[[177,172,168,166,160,157,161,153,145,147,149,134,142,144,149,155,154,148,148,141,146,145,147,145,14(...TRUNCATED) |
[[161,177,171,164,165,159,167,160,146,150,155,142,144,138,150,156,156,159,156,144,144,143,151,148,15(...TRUNCATED) |
[[168,170,167,169,167,160,162,169,156,144,141,146,148,141,149,151,157,161,154,141,146,145,144,145,15(...TRUNCATED) |
End of preview. Expand in Data Studio
NucMM-M Dataset (Mouse)
3D Neuronal Nuclei Segmentation in Mouse Brain Electron Microscopy.
Dataset Structure
├── Image/
│ ├── train/
│ │ ├── img_000_200_200.h5
│ │ ├── img_000_604_576.h5
│ │ ├── img_508_200_576.h5
│ │ └── img_508_604_200.h5
│ └── val/
│ ├── img_000_200_576.h5
│ ├── img_000_604_200.h5
│ ├── img_508_200_200.h5
│ └── img_508_604_576.h5
└── Label/
├── train/
│ └── seg_*.h5
└── val/
└── seg_*.h5
Label Information
- Original: Instance segmentation (each nucleus has unique ID)
- Classes: Background (0) + Nuclei instances (1, 2, 3, ...)
Related Datasets
- NucMM-Z (Zebrafish): Angelou0516/NucMM-Z
Usage
from huggingface_hub import hf_hub_download
import h5py
# Download a volume
local_path = hf_hub_download(
repo_id="Angelou0516/NucMM-M",
filename="Image/train/img_000_200_200.h5",
repo_type="dataset"
)
# Load with h5py
with h5py.File(local_path, 'r') as f:
volume = f[list(f.keys())[0]][:]
print(volume.shape) # (200, 200, 200)
Citation
If you use this dataset, please cite the NucMM paper:
@inproceedings{lin2021nucmm,
title={NucMM Dataset: 3D Neuronal Nuclei Instance Segmentation at Sub-Cubic Millimeter Scale},
author={Lin, Zudi and Wei, Donglai and Lichtman, Jeff and Bhanu, Bir and Bhattacharjee, Sumit and Bhavsar, Khushali and Bhosale, Pratik and Bhunre, Sandhya and others},
booktitle={MICCAI},
year={2021}
}
- Downloads last month
- 19