/MaskRCNN-Using-Detectron2-On-Custom-Dataset

This repository presents an object detection project using Mask R-CNN via Detectron2. A custom dataset of 10 dog and 10 cat images was created, annotated using Labelme, and resized to 600x800 pixels due to size mismatches. The model, trained on Google Colab, successfully detected and segmented cats and dogs in test images.

Primary LanguageJupyter Notebook

Watchers