Part 3: Developing a Custom Detectron2 Model for Instance Segmentation Tasks

The third part is another hands-on project. This part provides the readers with common tools for collecting and labeling images for object instance segmentation tasks. Additionally, it covers the steps to extract data from different sources and reconstruct a dataset in the format supported by Detectron2. Before training an object segmentation model, this section also utilizes the codes and visualizations approach to explain the architecture of an object segmentation application developed using Detectron2. The object instance segmentation models utilize results from the object detection models. Therefore, all the techniques introduced in the previous part for fine-tuning object detection models work the same for object instance segmentation models. However, object instance segmentation has an important feature to fine-tune: the quality of the boundaries of the detected objects. Therefore, this section also...