-
Notifications
You must be signed in to change notification settings - Fork 3
Getting Started: Running the Example
The repository includes two versions of example log files from a face detection algorithm's inference over several videos, as well as corresponding ground truth data for the same videos. These files serve as examples for users to run the Statistics Tool on their own machine learning models.
In addition, the example includes the Statistics Tool "external lib," which includes user-defined functions and report configurations necessary to run the example. Users can modify these configurations to fit their own needs and use them as a starting point for their own machine learning models.
To learn more about the configurations, user-defined functions, and report suite configurations included in the Statistics Tool "external lib," users can refer to the related wiki page. This page provides detailed information on how to modify and customize these configurations to fit their own needs.
1. Clone the repository into the desired folder.
2. Install python 3.10
3. Install the Python requirements by running the command "pip install -r requirements.txt"
1. Run the "statistic_tool.bat" batch file. This will set all the necessary parameters to run the example and point the "external_lib_path" parameter to the library under the "example1" folder.
2. Open a browser and type in the address "http://127.0.0.1:5000/".
3. You will be presented with the main page.
The example files include:
- Logs from two face detection model versions. - Ground truth log files. - Pre generated reports for each version, including two reports for each version: one for the classification scenario and one for the detection scenario - Preconfigured configurations for running new report suite
To run a new report from the example log files, follow these steps:
1. Navigate to the "New Report" page.
2. Choose a report suite or a preconfigured configuration shown in the browser.
3. Set the "Predictions Directory" to "StatisticsTool\running_example\source-logs\example-run-face-logs".
4. Set the "Ground Truth Directory" to "StatisticsTool\running_example\source-logs\example-run-annotations".
5. Set the "Output Directory" to a folder on your local machine.
6. Create the report.
To view the example report, follow these steps:
1. Navigate to the main page.
2. In the "Main Directory" field, insert "StatisticsTool\running_example\sample1\reports_output\example-run-suite-logs" as the main report directory.
3. (Optional - if you want to see comparing features) In the "Reference Directory" field, insert "StatisticsTool\running_example\sample1\reports_output\example-run-suite-ver2-logs".
4. Click on the "Load Report" button to load the report.