Advanced Marine Observations VALidation-Identification system based on Hybrid intelligence

A major bottleneck in AI is the need for significant human supervision and work to label training data and validate model outputs. As a result, many AI applications rely on the same training datasets and may not adequately examine outcomes. Hybrid Intelligence, which combines human and AI capabilities, could address the challenge of obtaining training sets and enable more effective learning.

To address this, ANERIS will develop AMOVALIH - a hybrid intelligent system for classifying marine life images, integrating reputation-based classification systems driven by human input with advanced automatic identification systems. These virtual agents will learn progressively from new validated observations of new species. This will be the first instance of a hybrid intelligence approach being developed for operational marine life monitoring.

Currently, the general API for generic AI integration with different AI’s is being implemented. Work is also being done on the frontend, considering best UX practices.

Target Stakeholders

Governmental Authorities: Fishery and aquaculture institutions 

Academia: Researchers & Citizen scientists

Governmental Authorities: Fishery and aquaculture institutions 

Contact / Technology lead:

Jaume Piera (

Partners involved