Positive, Negative, and False Results

When processing the images in the test set, the VisionPro Deep Learning application tries to predict whether the image belongs to a certain class, or whether it has a defect. The predictions of the tool can be either correct or incorrect, which are called positive and negative results.

• A positive result means that the tool has found what is was trained to find, such as a specific class, defect, or a feature A feature is a visually distinguishable area in an image. Features typically represent something of interest for the application (a defect, an object, a particular component of an object). in the image.

• A negative result means that the tool did not find what it was trained to find.

Positive and negative results can be incorrect, which are called false positive and false negative results:

• A false positive result occurs when the tool finds a class, defect, or feature that does not actually exist in the image. For example, the tool incorrectly identifies a scratch on a product as a defect when it is actually part of the design.

• A false negative result occurs when the tool does not identify a class, defect, or feature that exists in the image. For example, the tool fails to find a scratch on a damaged product.

When calculating statistical results for most tools, the VisionPro Deep Learning application compares the results from the tool with your labels, which represent the ground truth Ground truth refers to the actual nature of the problem that is the target of a machine learning model, reflected by the relevant data sets associated with the use case in question. Supervised machine learning models are trained on labeled data that are considered for the model to identify patterns that predict those labels on new data. for the tool. Then, the application uses the rate of false positives and false negatives to calculate the Precision and Recall scores.