---
license: mit
language:
- en
metrics:
- f1
- precision
- recall
- mape
pipeline_tag: object-detection
tags:
- biology
- climate
---


The model was trained using Ultralytics Yolov8m model 

The dataset was the ReWilding Europe dataset which has around 70+ species, however a smaller sample of 30 classes was taken. 
The criteria for selection was the data distribution with a minimum cutoff for each class. 

---
A European Fallow Deer. 
![image](fallow_deer.jpg)
---

Model performance
![image](confusion_matrix_normalized.png)
![image](F1_curve.png)

---
co2_eq_emissions:

  emissions: 9.07 kg of CO2 eqv (in grams of CO2)
  
  source: "source of the information, directly from code carbon or from a scientific article documenting the model"
  
  training_type: "pre-training or fine-tuning" for 60 hours (aprox)
  
  geographical_location: Western India
  
  hardware_used: 1 RTX 3090
![image](CarbonCalculations.png)
  
---