Before continuing, make sure you have your desired models uploaded to the server in Model Management Page.
In Model Selection and Configuration, you will need to select the model containing objects of interest.
Object Detection Model
- Select the object detection models you want to add from the model drop-down list. All object classes from the model will be displayed after selected
- Adjust the min score value (if needed)
- Finally, click Add to add the model to the task group
Built-in License Plate Detection Model
- Select the License Plate Detection model from the model drop-down list. There is no extra configuration required for this model
- Finally, click Add to add the model to task group
Built-in Facial Recognition Model
- Select the Facial Recognition model from the model drop-down list. There is no extra configuration required for this model
- Finally, click Add to add the model to task group
Built-in Face Detection Model
- Select the Face Detection model from the model drop-down list. There is no extra configuration required for this model
- Finally, click Add to add the model to task group
Built-in OCR Model
- Select the OCR model from the model drop-down list.
- Select the target Language you want to detect
- GPU ID: You can select a different NVIDIA GPU (if the PC has more than 1 GPU card)
- Detector threshold: the threshold score for the OCR detector that finds the bounding boxes around texts
- Detector BBox threshold: the IoU score for Bounding box detection
-
Detector unclip ratio: The larger the unclip ratio, the larger the range of the text detection box; the smaller the value, the closer the text box is to the text.
-
Classifier threshold: Text classifier score
-
Use Dilation: Add dilation algorithm for text classifier
- Finally, click Add to add the model to task group