An object detection model to count birds nesting. The model will place a box around each bird detected, allowing extraction of metrics about the size and location of the bird. Optionally we try to classify the bird species.
- Using Yolo, with dataset: Fast, Deep Detection and Tracking of Birds & Nests
- Using Retinanet: An Automatic Detection Method of Bird’s Nest on Transmission Line Tower Based on Faster_RCNN
- Predicting Bounding Boxes using Cornell’s NABirds Data. with EfficientNet uses the NABirds V1 collection of 48,000 annotated photographs of the 400 species of birds that are commonly observed in North America
- Dataset 250 Bird Species 35215 Train, 1250 Test, 1250 Validation images 224X224X3 jpg format but no bounding boxes, see example notebook
- Dataset 200 Bird Species with 11,788 Images with bounding boxes
- Dataset Interspecies classification of species in High Resolution Images but no bounding boxes