Classify pixel using deep learning

sharmaakki

Dear All

I have been trying to use U-Net to train my deep learning model for classification of satellite data pixels but while train deep learning  model stage while selecting model parameter "U-Net" it shows an "ERROR 000800 The value is not a member of Single Shot Detector (Object detection) | RetinaNet (Object detection) | YOLOv3 (Object detection) | FasterRCNN (Object detection)."

I am using ArcGIS pro desktop interface.

I am not able to figure out why it is showing only object detection models not pixel detection models?

Please help me out in this.

 

Thanks in advance.

 

Anonymous User

Hi,

In the previous step you have probably exported the training data in Pascal VOC Format. For training a Unet (Pixel Classification) Model, you need to export the training data in Classified Tiles format.

Thanks,

Sandeep