Name:
COMPOUND.AI
Developed by:
Marine Technology Centre (CTN)
Countrie(s) involved:
SPAIN
Funding sources:
National (ministries, agencies, regions)
Specific organism(s) focused:
Platform for the automation of the deployment, diagnosis and maintenance of artificial vision solutions based on Artificial Intelligence
The general objective of the project is to develop a platform that facilitates and automates the deployment of artificial vision solutions based on Artificial Intelligence, overcoming the challenges of scalability and outdated biases, evaluating it through application to a use case in the marine environment, taking advantage of different technologies such as edge computing, cloud computing, federated learning, or MLops, to avoid the appearance of storage, processing and communication bottlenecks, and allowing easy scaling of applications.
To improve the identification and classification processes of marine species (fauna and flora), a variety of tools will be implemented that will strengthen artificial vision in the marine environment. COMPOUND.AI will not only open up new opportunities for marine environment monitoring activities, but will also drive the advancement of technology.
Benefits for the marine environment: Help in monitoring species and ecosystems for their conservation. Extract methodologies to improve decision-making regarding the marine environment
Benefits for the industry: Counting individuals (aquaculture). Monitoring species in the external environment of aquaculture facilities
Benefits for the scientific community: Improve knowledge about wireless communications at sea. Advance the state of the art on autonomous artificial vision platforms at sea.
Innovative aspect: COMPOUND.AI will allow the autonomy and self-assessment of the models, in which a self-assessment logic will be implemented within the system, so that it is autonomously capable of detecting when it is not working correctly and correcting itself by using other models available on the server or by sending new images for fine-tuning.
Specific organism(s) focused:
No
Serctor(s) involved:
No
TRL:
TRL 4-6
Helix sector/involvement:
Research performing organizations and academia; Ministries, agencies, local administrationYes
Link to the good practice:
https://ctnaval.com/proyectos/compoundai/