PacificCommunity/cfap-ai-models

Deep learning models created for analysis fish and invertebrates pictures from market and landing surveys

SPC CFAP Neural Network Models

Introduction

The Coastal Fisheries and Aquaculture Programme of the Pacific Community develops neural network models for the analysis of fish and invertebrates pictures taken as part of market & landing surveys.

This repository contains the main models that are used as part of a image processing chain to calibrate images, detect, measure and identify specimens. Most detection models are based on YOLO v4, but we also use Detectron 2 PointRend for contouring and ResNet-101 for classifiers.

Models

• YOLO v4 - Survey photo types
• YOLO v4 - SPC Board 3 detection
• YOLO v4 - SPC Mat 1 detection
• YOLO v4 - Fishes on SPC Mat 1 detection

License

This work is licensed under Creative Commons Attribution-Non Commercial 4.0 license (CC BY-NC 4.0)

Acknowledgements

This work has been funded by the Australian Department of Foreign Affairs and Trade (DFAT), European Union through Pacific-European Union (EU) Marine Partnership (PEUMP) programme and Australian Center for International Agricultural Research (ACIAR) through Australian National Centre for Ocean Resources and Security (ANCORS) Pathways 2 project