Periodic Reporting for period 1 - iMagine (Imaging data and services for aquatic science)
Période du rapport: 2022-09-01 au 2023-08-31
iMagine offers AI-powered image analysis tools and marine and freshwater research datasets, promoting more efficient data processing. This framework, integrated into the European Open Science Cloud, supports AI model development, benefiting various applications such as water pollution mitigation, biodiversity studies and others.
The project leverages the EGI federation infrastructure, providing neural networks and distributed data analysis capabilities. Over 9 million images and 8 AI applications from 13 Research Infrastructures are shared via this framework, fostering Best Practices development. Synergies among aquatic use cases facilitate shared solutions in data management, quality control, and FAIRness, contributing to harmonization across RIs and Best Practice guidelines.
The project leverages the EGI federation infrastructure, providing neural networks and distributed data analysis capabilities. Over 9 million images and 8 AI applications from 13 Research Infrastructures are shared via this framework, fostering Best Practices development. Synergies among aquatic use cases facilitate shared solutions in data management, quality control, and FAIRness, contributing to harmonization across RIs and Best Practice guidelines.
In the first year of iMagine, significant advancements were made in AI-powered image analytics for marine and freshwater research. Key achievements and activities are summarised:
Objective 1: Deliver a scalable, shared IT platform for image analysis in marine and freshwater research.
iMagine AI Platform launched in December 2022, based on the DEEP Platform. Online webinars and face-to-face tutorials were conducted for consortium members. Integration with CSIC and LIP cloud compute sites provided substantial computational resources. Successful federation of the Walton OpenStack cloud into EGI's federation. The platform was upgraded to AI4EOSC technology in Q2 2023, enabling external use cases.
Objective 2: Advance existing image analytical services to increase research performance in aquatic sciences. Five mature use cases advanced towards production-level image analytics services. Work focused on labelling training images, model migration, and improving prediction accuracy. Some use cases integrated models with image pre-processing services, achieving notable results like the fish classification application by OBSEA-EMSO.
Objective 3: Develop prototype new image analytical services and datasets that can accelerate progress towards healthy oceans, seas, coastal and inland waters. Three prototype use cases, new to AI-based image processing, achieved promising results in object classification with high accuracy.
Objective 4: Capture and disseminate development and operational best practices to imaging data and image analysis service providers. Both platform providers and use case developers documented Good practices. The project initiated a partnership with the AI4Life project.
Objective 5: Deliver a portfolio of scientific image and image analytics services targeting researchers in marine and aquatic sciences. Mature use cases progressed in development, with the first expected to serve external users in late 2023. Services include trained models for image classification and FAIR images used during model training.
In summary, iMagine has made significant progress in infrastructure development, improving existing services, developing new solutions, and sharing best practices. The project is on course to provide valuable image analysis services for marine and aquatic researchers.
Objective 1: Deliver a scalable, shared IT platform for image analysis in marine and freshwater research.
iMagine AI Platform launched in December 2022, based on the DEEP Platform. Online webinars and face-to-face tutorials were conducted for consortium members. Integration with CSIC and LIP cloud compute sites provided substantial computational resources. Successful federation of the Walton OpenStack cloud into EGI's federation. The platform was upgraded to AI4EOSC technology in Q2 2023, enabling external use cases.
Objective 2: Advance existing image analytical services to increase research performance in aquatic sciences. Five mature use cases advanced towards production-level image analytics services. Work focused on labelling training images, model migration, and improving prediction accuracy. Some use cases integrated models with image pre-processing services, achieving notable results like the fish classification application by OBSEA-EMSO.
Objective 3: Develop prototype new image analytical services and datasets that can accelerate progress towards healthy oceans, seas, coastal and inland waters. Three prototype use cases, new to AI-based image processing, achieved promising results in object classification with high accuracy.
Objective 4: Capture and disseminate development and operational best practices to imaging data and image analysis service providers. Both platform providers and use case developers documented Good practices. The project initiated a partnership with the AI4Life project.
Objective 5: Deliver a portfolio of scientific image and image analytics services targeting researchers in marine and aquatic sciences. Mature use cases progressed in development, with the first expected to serve external users in late 2023. Services include trained models for image classification and FAIR images used during model training.
In summary, iMagine has made significant progress in infrastructure development, improving existing services, developing new solutions, and sharing best practices. The project is on course to provide valuable image analysis services for marine and aquatic researchers.
The iMagine AI platform is now accessible via the EOSC Marketplace and soon AIoD. Project's Key Exploitable Results are linked to each use case, showcasing their progress and expected impact:
UC1 "Marine litter assessment" refined processing methodology, model development, GUI integration, and mapping to specific categories. Impact: Important for environmental management, cleaning operations, and supporting the EU Marine Strategy Framework Directive and EU Green Deal.
UC2 "Zooscan - EcoTaxa pipeline" advanced the AI-powered identification pipeline for plankton. Impact: Essential for ecosystem services, MSFD, and WFD descriptors. Aids in understanding food availability and climate change effects.
UC3-OBSEA achieved promising results in fish classification and is building a higher-resolution training dataset.
UC3-Azores improved training data quality for species detection, achieving over 90% accuracy.
UC3-Smartbay explored annotation and Quality Control, contributing to biodiversity and ecosystem studies.
UC4 "Oil spill detection" established a framework for oil spill forecasting and curated a Mediterranean Sea dataset. Impact: Enhancing oil spill monitoring for professional users.
UC5 "Flowcam plankton identification" optimized data pipelines, achieved high accuracy in phytoplankton identification, vital for Good Environmental Status assessment.
These prototype services contribute to marine ecosystem preservation, efficient environmental monitoring, and addressing environmental challenges across Europe. Additional progress was made in underwater noise spectrograms, beach monitoring, and freshwater diatoms identification.
Two additional KERs need to be mentioned:
The iMagine AI Platform, whose expected impact is to allow users to rapidly develop, iterate and optimise AI models and eventually serve them to end users.
iMagine Best Practices will be a versatile resource for projects, researchers, students, and other stakeholders alike, enabling them to delve into AI techniques and begin their explorations. This knowledge repository will guide future AI adopters and users, providing a roadmap for harnessing the potential of AI within the aquatic sciences and beyond
UC1 "Marine litter assessment" refined processing methodology, model development, GUI integration, and mapping to specific categories. Impact: Important for environmental management, cleaning operations, and supporting the EU Marine Strategy Framework Directive and EU Green Deal.
UC2 "Zooscan - EcoTaxa pipeline" advanced the AI-powered identification pipeline for plankton. Impact: Essential for ecosystem services, MSFD, and WFD descriptors. Aids in understanding food availability and climate change effects.
UC3-OBSEA achieved promising results in fish classification and is building a higher-resolution training dataset.
UC3-Azores improved training data quality for species detection, achieving over 90% accuracy.
UC3-Smartbay explored annotation and Quality Control, contributing to biodiversity and ecosystem studies.
UC4 "Oil spill detection" established a framework for oil spill forecasting and curated a Mediterranean Sea dataset. Impact: Enhancing oil spill monitoring for professional users.
UC5 "Flowcam plankton identification" optimized data pipelines, achieved high accuracy in phytoplankton identification, vital for Good Environmental Status assessment.
These prototype services contribute to marine ecosystem preservation, efficient environmental monitoring, and addressing environmental challenges across Europe. Additional progress was made in underwater noise spectrograms, beach monitoring, and freshwater diatoms identification.
Two additional KERs need to be mentioned:
The iMagine AI Platform, whose expected impact is to allow users to rapidly develop, iterate and optimise AI models and eventually serve them to end users.
iMagine Best Practices will be a versatile resource for projects, researchers, students, and other stakeholders alike, enabling them to delve into AI techniques and begin their explorations. This knowledge repository will guide future AI adopters and users, providing a roadmap for harnessing the potential of AI within the aquatic sciences and beyond