## Dataset Information
This dataset focuses on characterizing crystallites within two types of crystals, Spherulite and Lamellar, using Atomic Force Microscopy (AFM) images acquired at Lawrence Livermore National Laboratory (LLNL).
Currently, the dataset consists of ~500 annotated images.
- Images and corresponding ground truth masks can be found [here][1].
**Metadata:**
- Imaging technique: AFM
- Sample: Crystallites
- Types: Spherulite and Lamellar
- Image size: 512 x 512
- Resolution: High-resolution
**Want to contribute? Fill up this [Google Form][2]**.
<h2>
Benchmarks
</h2>
| Model | Description | Binary IoU | Library | Input Image Size | Performance Metrics | Paper | Evaluation Script |
|:---------------------------------:|:--------------------------------------------:|:----------:|:--------------------:|:----------------:|:------------------------------------------:|:------------------------------------------:|:------------------------------------------:|
[trial_63.ckpt][3] | U-Net++ with Xception Encoder Backbone | 72% | PyTorch <br> ([Segmentation Models][4]) | 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU* | - | - |
|[trial_84.ckpt][5]| U-Net model Xception Encoder Backbone| 72% |PyTorch <br> ([Segmentation Models][6])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_8.ckpt][7]| U-Net model Xception Encoder Backbone| 69% |PyTorch <br> ([Segmentation Models][8])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_15.ckpt][9]| U-Net++ model Xception Encoder Backbone| 68% |PyTorch <br> ([Segmentation Models][10])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_5.ckpt][11]| U-Net++ model Xception Encoder Backbone| 67% |PyTorch <br> ([Segmentation Models][12])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_112.ckpt][13]| U-Net++ model Xception Encoder Backbone| 66% |PyTorch <br> ([Segmentation Models][14])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_22.ckpt][15]| U-Net model Xception Encoder Backbone| 64% |PyTorch <br> ([Segmentation Models][16])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_58.ckpt][17]| U-Net++ model SE-ResNext101 Encoder Backbone| 62% |PyTorch <br> ([Segmentation Models][18])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_106.ckpt][19]| U-Net model SE-ResNext101 Encoder Backbone| 62% |PyTorch <br> ([Segmentation Models][20])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_108.ckpt][21]| U-Net model SE-ResNext101 Encoder Backbone| 62% |PyTorch <br> ([Segmentation Models][22])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_76.ckpt][23]| U-Net model SE-ResNext101 Encoder Backbone| 61% |PyTorch <br> ([Segmentation Models][24])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
|[trial_11.ckpt][25]| U-Net++ model SE-ResNext101 Encoder Backbone| 59% |PyTorch <br> ([Segmentation Models][26])| 512 x 512| Accuracy <br> Precision <br> Recall <br> Binary IoU*| - | - |
[1]: https://osf.io/b2wdg/files/osfstorage
[2]: https://forms.gle/4pGRhkdw4wGNy1dY9
[3]: https://osf.io/download/6659f4fa6b6c8e20c404cb9e/
[4]: https://github.com/qubvel/segmentation_models
[5]: https://osf.io/download/6670368965e1de5d78893c5a/
[6]: https://github.com/qubvel/segmentation_models
[7]: https://osf.io/download/6670321c6b6c8e331404cab8/
[8]: https://github.com/qubvel/segmentation_models
[9]: https://osf.io/download/6670336bd835c4389b4ce005/
[10]: https://github.com/qubvel/segmentation_models
[11]: https://osf.io/download/667032426b6c8e331404cac1/
[12]: https://github.com/qubvel/segmentation_models
[13]: https://osf.io/download/667037ed6b6c8e332204cb03/
[14]: https://github.com/qubvel/segmentation_models
[15]: https://osf.io/download/6670347177ff4c5e1ee044b0/
[16]: https://github.com/qubvel/segmentation_models
[17]: https://osf.io/download/6670364d6b6c8e331d04ceb8/
[18]: https://github.com/qubvel/segmentation_models
[19]: https://osf.io/download/66703757d835c438a44cdfa2/
[20]: https://github.com/qubvel/segmentation_models
[21]: https://osf.io/download/667037e50f8c8016f03c93fa/
[22]: https://github.com/qubvel/segmentation_models
[23]: https://osf.io/download/667036020f8c8016e63c9335/
[24]: https://github.com/qubvel/segmentation_models
[25]: https://osf.io/download/667034860f8c8016eb3c90d2/
[26]: https://github.com/qubvel/segmentation_models
