🔬 FluoGen: AI-Powered Fluorescence Microscopy Suite
Paper: FluoGen: An Open-Source Generative Foundation Model for Fluorescence Microscopy Image Enhancement and Analysis
Homepage: Homepage Website
Code: GitHub Repository
Select a task below.
Note: The "Pseudocolor" option instantly applies to both Input and Output images for better comparison. Use the "Download Raw Output" button to get the scientific 16-bit/Float data.
Instructions
- Select a desired prompt from the dropdown menu.
- Adjust the 'Inference Steps' slider to control generation quality.
- Click the 'Generate' button to create a new image.
- Explore the 'Examples' gallery; clicking an image will load its prompt.
Notice: This model currently supports 3566 prompt categories. However, data for many cell structures and lines is still lacking. We welcome data source contributions to improve the model.
Instructions
- Upload a low-resolution 9-channel TIF stack, or select one from the examples.
- Select a 'Super-Resolution Model' from the dropdown.
- Enter a descriptive 'Prompt' related to the image content (e.g., 'CCPs of COS-7').
- Adjust 'Inference Steps' and 'Seed' as needed.
- Click 'Generate Super-Resolution' to process the image.
Notice: This model was trained on the BioSR dataset. If your data's characteristics differ significantly, please consider fine-tuning the model using our project on GitHub for optimal results.
Instructions
- Upload a noisy single-channel image, or select one from the examples.
- Select the 'Image Type' from the dropdown to provide context for the model.
- Adjust 'Inference Steps' and 'Seed' as needed.
- Click 'Denoise Image' to reduce the noise.
Notice: This model was trained on the FMD dataset. If your data's characteristics differ significantly, please consider fine-tuning the model using our project on GitHub for optimal results.
Instructions
- Upload a single-channel segmentation mask (
.tiffile), or select one from the examples gallery below. - Enter the corresponding 'Cell Type' (e.g., 'CoNSS', 'HeLa') to create the prompt.
- Select how many sample images you want to generate.
- Adjust 'Inference Steps' and 'Seed' as needed.
- Click 'Generate Training Samples' to start the process.
- The 'Generated Samples' will appear in the main gallery, with the 'Input Mask' shown below for reference.
Notice: This model was trained on the 2018 Data Science Bowl dataset. If your data's characteristics differ significantly, please consider fine-tuning the model using our project on GitHub for optimal results.
Instructions
- Upload a single-channel image for segmentation, or select one from the examples.
- Select a 'Segmentation Model' from the dropdown menu.
- Set the expected 'Diameter' of the cells in pixels. Set to 0 to let the model automatically estimate it.
- Adjust 'Flow Threshold' and 'Cell Probability Threshold' for finer control.
- Click 'Segment Cells'. The result will be shown as a dark red overlay on the original image.
Notice: This model was trained on the 2018 Data Science Bowl dataset. If your data's characteristics differ significantly, please consider fine-tuning the model using our project on GitHub for optimal results.
Instructions
- Upload a single-channel image for classification, or select an example.
- Select a pre-trained 'Classification Model' from the dropdown menu.
- Click 'Classify Image' to view the prediction probabilities for each class.
Note: The models provided are ResNet50 trained on the 2D HeLa dataset.