Image Deep Learning Custom Training

Image deep learning custom training function: can use annotated data to train deep learning models, and extract information from images through model reasoning, including road sign detection (target detection), lane segmentation (semantic segmentation), road sign extraction, etc. Use trained models to automatically process large amounts of data.

This feature supports two business processes, namely:

(1)Training sample production -> Training model

(2)Use existing models to process image data

• Label image data using data labeling tools
• Training image deep learning models
• Training report
• Detect or segment objects with a trained model
• 3D Mapping
• Object detection training case tutorial
• Semantic segmentation training case tutorial
• Road marking training case tutorial