Get more targeted results with IKOSA AI by using custom training parameters.
Sometimes the parameter values predefined in IKOSA do not exactly match your research design. In such a case you need to be able to change parameter values to be able to answer your research question more precisely.
What to expect from this page?
📑 What training parameters can be customized?
The application training has been created for particular life science use cases. When training deep learning models, preset parameter values can be limiting when it comes to more specific image analysis tasks (e.g. various scales of objects or very large objects). IKOSA allows you to override default training parameter values and set custom ones for a more tailored training experience.
Training parameters
Below we provide training parameters whose values can be customized.
downscaleFactor
The downscale factor has a positive integer value between 1 and 1,000 (1, 2, 3, …, 1,000) determining the resolution at which the images are used during training. A higher value means that the images will be downscaled to a smaller resolution before being fed into the model (for example, with a downscale factor of 2, images will be downscaled to half of their original size).
A higher downscale factor can result in faster training and faster prediction when using the deployed IKOSA AI app.
For small objects, accuracy may decrease when a higher downscale factor is applied. However, in some cases, when large and weakly textured objects that exceed a radius of 32 pixels need to be identified, the model’s accuracy can also improve significantly.
The optimal choice for the downscale factor strongly depends on the use-case and this parameter needs to be tuned experimentally.The default downscale factor values used for the different training types are as follows:
Semantic Segmentation: 2
Instance Segmentation (2D): 1
Instance Segmentation (Multichannel): 1
Please note: These training parameters values are automatically set up in the deployed IKOSA AI applications.
📑 How to set custom values for the training parameters?
Custom values for training parameters can be defined as:
Via IKOSA AI Wizard (NO technical knowledge required)
Via API (technical knowledge required)
How to set them up via IKOSA AI Wizard?
In the last step of your training session setup, you can select the downscale factor from the "Advanced settings" drop-down section.
How to set them up via API?
When starting an IKOSA AI training via an IKOSA Prisma API call (POST request to /training endpoint, see https://prisma.ikosa.ai/api/v1/docs/#/command/createTraining).
Please note: Using custom values for training parameters requires technical knowledge regarding HTTP requests via API.
Here is how starting training with custom parameter values looks like:
Define your data for the JSON payload of the POST request to the
/trainingendpoint and define custom training parameter values inhyperparamOverrides.
For example:{ "data": { ..., "hyperparamPreset": "QUICK", "hyperparamOverrides": { "downscaleFactor": 4 } } }In the header of the request, include the appropriate authentication information, such as an API key or token.
Send the POST request to the API endpoint.
The API will then process the request and return a response, including information about the training location or about errors.
The created training is shown in your IKOSA account under IKOSA AI > Training Overview.
You are now good to go and create your first IKOSA AI application training with custom training parameter values. You will see immediate improvements to the performance of your app!
If you need any help evaluating the performance of your app, feel free to send us an email at support@ikosa.ai. Copy-paste your training ID into the subject line of your email.
Share this tip! Help others to improve their research outcomes with the help of custom parameter values.
If you still have questions regarding your application training, feel free to send us an email at support@ikosa.ai. Copy-paste your training ID into the subject line of your email.