Deep Learning Model
- How to determine if a scenario is suitable for using the deep learning models?
-
As long as the on-site objects contain features similar to the objects that the deep learning models can recognize, the models can be deployed for testing.
-
If the on-site objects cannot be recognized, you should train a new model.
-
If some objects are correctly recognized, you should acquire images that report poor recognition results and fine-tune the deep learning models.
-
- What scenarios can the deep learning models for cartons be applied to?
-
It is suitable for palletizing/depalletizing cartons of single or multiple colors and surface patterns. However, please note that this deep learning model is only applicable to cartons placed in horizontal layers and not at an angle to the ground.
- How to acquire data for the deep learning model for cartons?
-
Please test the deep learning model first. If it cannot segment correctly sometimes, acquire about 20 images of situations where the model does not perform well and train the deep learning model.