StomataPy is a Collection of Resources that Enable Generalized Stomatal Segmentation via Community-Driven Human-in-the-Loop
Namely, our resources allow the stomatal community to contribute images together, to iterativly improve open source models
- StomataPy400K: The largest annotated stomata dataset ever (Opensource soon)
- ISAT-SAM: Interactive Stomata Annotation Tool with Segment Anything Model (Opensourced already)
- StomataPy400K models: a series of models for stomata related segmentations (Opensource soon)
- Total images: 7,838
- Total plant species: 425
- Total images with masks: 393,671 (Autolabeled: 290,898, 73.9 %)
├── Superclasses
├── 'pavement cell': 113,561
├── 'stomatal complex': 168,084
├── Subclasses of 'stomatal complex'
├── 'stoma': 97,691
├── 'outer ledge': 11,928
├── 'pore': 2,407
- Total_modalities: 7
├── ClearStain_Brightfield
├── Imprints_Brightfield
├── Imprints_DIC
├── Leaf_Brightfield
├── Leaf_Topometry
├── Peels_Brightfield
├── Peels_SEM
The dataset will be shared on HuggingFace: https://huggingface.co/datasets/aliasz/StomataPy400K
Already available on GitHub: https://github.com/yatengLG/ISAT_with_segment_anything
The models will be shared on HuggingFace: https://huggingface.co/aliasz/StomataPy400K-Models
Note: you need the secret key to access the models. If you are interested in testing the models, please contact me at [email protected]
We greatly appreciate the following beta-test participants:
Sara Paola Nastasi and Alex Costa from University of Milan, Italy
Robert Caine, Nitkamon Iamprasertkun, Yixiang Shan, and Safia El Amiri, from University of Sheffield, UK
Ron Eric Stein and Tabea Lara Zwaller, from Universität Heidelberg, Germany
Didier Le Thiec Lab from INRAE, France
Emilio Petrone Mendoza from University of Naples Federico II, Italy
Hana Horak from University of Tartu, Estonia
Mengjie Fan and Tracy Lawson, from University of Essex, UK
Pawandeep Singh Kohli and Micheal Rasissig from Uiversity of Bern, Switzerland
Nattiwong Pankasem from University of California San Diego, USA
Xiaojuan Wang from Shanghai Science & Technology Museum, China
Xiaoqian Sha and Tian Zhang from Henan University, China