Software Requirements/Operations
- Whether AMD CPUs can run CPU models?
-
AMD CPUs do not support running CPU models.
- Why does the inconsistency between the ROI settings of on-site data and training data affect the confidence values of instance segmentation?
-
The inconsistency will result in objects being out of the optimal recognition range of the model, thus affecting the confidence. Therefore, please keep the ROI settings of the on-site data and training data consistent.
- How to set a proper ROI?
-
-
For objects carried by trays and bins, the ROI should cover the carriers in all images, and a margin of about one-third of the carrier size should be kept, which can reduce picking deviations due to the worn carrier or shaking during picking.
-
For objects carried by conveyor belts, the ROI should closely fit the edges of the objects.
-
The ROI should not cover objects unrelated to the target objects.
-
If the ROI cannot suit all scenarios, you can preprocess images in Mech-Vision by scaling images in 2D ROI, setting 3D ROI, and calculating color images for the highest layer.
-
- ROI position deviations may occur when opening old projects with the newer-version Mech-DLK.
-
The ROI will be corrected after you click Validate.
- When training a model in Mech-DLK, what should I do if the error message “ModuleNotFoundError: No module named ‘onnxruntime’” shows?
-
Go to the “users” folder in OS (C:) and open the folder of the current user of the computer. Check if the folder “AppData/Roaming/Python/Pythong36/site-packages” is empty. If not, please delete all contents in the folder.
- How to improve labeling efficiency?
-
-
Mech-DLK provides the smart labeling suite to improve labeling efficiency. You can combine the tools to meet on-site accuracy and cycle time requirements.
-
Use the VFM Labeling Tool to label data in batches and validate the models to see if the on-site accuracy and cycle time requirements are met.
-
If not, it is recommended to use the VFM Labeling Tool in combination with the Smart Labeling Tool to manually fine-tune the data until the requirements are met.
-
-
Mech-DLK allows you to import and export datasets. You can export unlabeled data and distribute it to multiple labelers.
-
- How to use multiple labeling results for model training?
-
Export the labeled data completed by each labeler separately, and then import all exported datasets into the same project. After the datasets are imported, check if the number of images has increased by the same amount as the number of imported images. If they match, it indicates a successful import, and you can use the imported data for model training and other operations.