For my MEng thesis at Imperial College London, I am building deep learning model that automatically postprocesses shark BRUVs by detecting species, behaviours and metadata such as size.
This would streamline shark conservation efforts, eliminating the costly and error-prone manual processing.
The accuracy of the model will ultimately depend on the quality of data used for training. Therefore, I am looking for scientists and organisations that work with Shark BRUVs and would be willing to share their annotated videos.
The model is open source and I would be happy to contribute back by adapting it to your specific Shark-related use cases!
Thank you very much