Case Study

Campus-Objects - Object Detection using LW-DETR

Object detection project using LW-DETR and PyTorch on 3,000+ annotated campus images, achieving 0.71 mAP@0.5.

PyTorchLW-DETRObject DetectionCOCOComputer Vision

Project summary

Campus-Objects is a multi-class object detection project built around annotated campus imagery and LW-DETR.

Dataset

The project uses 3,000+ annotated campus images prepared for object detection workflows.

Technical approach

The pipeline uses PyTorch and LW-DETR for object detection, with dataset preparation aligned to COCO-style training workflows.

Evaluation and results

The model achieved 0.71 mAP@0.5 on the available evaluation setup.

Challenges and limitations

The repository content does not provide enough detail to state deployment status, production usage, or per-class performance.