Final Bee Health Model (V3) Computer Vision Project
Updated 8 months ago
Metrics
This is the final iteration of the bee health assessment model I created. Using 3 classes in image segmentation (bee, pollen, and mite) it can determine the overall health variables of a local bee population. The data used includes examples of both male and female Western Honeybees (Apis mellifera). This species is common in the United States, although it is important to note small variation in the physiological features between sexes could make the model less accurate. It is recommended to use isolated, close up images of bees for best accuracy. This model was created for a class assignment in AI and Natural History at St. Mary’s College of Maryland.
Images were taken using research grade images from INaturalist and VarroaDataset from The Computer Vision Lab TU Wien.
Use This Trained Model
Try it in your browser, or deploy via our Hosted Inference API and other deployment methods.
Build Computer Vision Applications Faster with Supervision
Visualize and process your model results with our reusable computer vision tools.
Cite This Project
If you use this dataset in a research paper, please cite it using the following BibTeX:
@misc{
final-bee-health-model-v3_dataset,
title = { Final Bee Health Model (V3) Dataset },
type = { Open Source Dataset },
author = { Erin Tomassoni Workspace },
howpublished = { \url{ https://universe.roboflow.com/erin-tomassoni-workspace/final-bee-health-model-v3 } },
url = { https://universe.roboflow.com/erin-tomassoni-workspace/final-bee-health-model-v3 },
journal = { Roboflow Universe },
publisher = { Roboflow },
year = { 2024 },
month = { apr },
note = { visited on 2024-12-23 },
}