Bird Sound Classifier on the Edge

We performed live classification of data using both our smartphones as well as the Arduino Nano 33 BLE Sense. Every Edge Impulse project has a test dataset in addition to its training data. The test dataset is immediately saved with the samples taken in Live classification, and the Model testing page displays all of the test data. To use the sample that was captured for testing, the expected outcome should be edited accordingly. Click the icon and select Edit expected outcome, then enter the relevant label, as shown below. Now, select the sample using the checkbox to the left of the table and click Classify selected. We can observe that the model's accuracy has been rated based on the test data. As expected, the performance of the model isn't always great on the first attempt, which can be so due to several factors.