Arctic Seal Detection

Monitoring the count of any animal is important for understanding its habitat and helps in its conservation. The count can give insights about the animal's habitat, breeding site, migration pattern, and other behaviors. Seals form an important part of the food chain in the cryosphere region of the Arctic. The effects of global warming and climate change can impact sea ice and, thus, the seals. Vulnerable animals such as polar bears feed on these seals. Reduction in the seal population can heavily impact the ecosystem. 

Developments in drone imagery and associated technology have led to advancements in detection and identification of different types of wildlife. Arctic Seals live in extreme climatic conditions that are uninhabitable to humans. Drones can work in these extreme conditions and cover vast areas in less time. Deep learning models applied on drone imagery can be effective in detecting animals and can help in census counts. This model detects ringed seals and brown bearded seals in high-resolution drone imagery of the Arctic region.

Using the model
Follow the guide to use the model. Before using this model, ensure that the supported deep learning libraries are installed. For more details, check Deep Learning Libraries Installer for ArcGIS.

Fine-tuning the model
This model can be fine-tuned using the Train Deep Learning Model tool. Follow the guide to fine-tune this model.

Input
High resolution RGB imagery (3 - 5 centimeter spatial resolution). 

Output
Feature class containing detected seals.

Applicable geographies
The model is expected to work well in cryosphere region of Arctic.

Model architecture
This model uses the FasterRCNN model architecture implemented in ArcGIS API for Python.

Accuracy metrics
This model has an average precision score of 0.77 for seals with F-1 score of 0.87 .

Training data
The model has been trained on the NOAA Arctic Seals 2019 dataset.

Limitations