Deepsense is a hybrid deep learning system designed for the automated detection and classification of railway defects. By combining the speed of YOLOv8 with the pixel-level accuracy of Mask R-CNN, Deepsense ensures high-precision defect localization in real-time. The system is integrated into a local web-based interface, allowing field personnel to receive immediate feedback from camera feeds without the need for cloud infrastructure.
erirem/deep-track
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|