Artificial intelligence (AI) is a useful tool for analyzing data and supporting research. While the currently more popular forms of generative AI is impacting our daily lives and society and carries a considerable amount of risks, the intergration of predictive AI in interactive Apps can be highly beneficial for data analysis and science applications.
On SPOTTERON, we already utilize AI for identifying invasive species. Thanks to the Horizon 2020-funded project "IPM-Popillia", we have developed an integrated AI component for Citizen Science Apps that enables users to receive suggestions on the animal species depicted in their photographs, alongside a confidence score for each suggested species. Previously launched as a prototype integration, the Citizen Science App with AI Species ID is out now as a live innovation demonstrator outcome from the EU Horizon Europe programme.
The SPOTTERON AI System is developed and published in various stages. From the prototype integration in phase 1 we extend the functionality to utilize not only one uplaoded image but all from one user uploaded spot from the Multiple Images Upload feature launched in SPOTTERON FOUR. Further phases enable the App users to highlight poins of interest in their pictures, which is a helpful tool if for example multiple animal and plant species are present in the same photographs like on flowers with pollinators and bee species visiting.
The AI tool in the Citizen Science App of the project goes even further. We have integrated the SPOTTERON Taxonomy System in the user interface dialog. This enables to manually enter a species name independent from the outcome of the AI image recognition process with a world wide taxonomy data base in the background, including common names in many languages and variations, scientific names and even species information as a glossary and e-learning pages. Here is more information about the SPOTTERON Taxonomy System, available for all Citizen Science Apps on the platform: Citizen Science App Taxonomy Service Info Page
In Citizen Science, topics for projects come from a wide range of areas. From biodiversity monitoring to recording habitat developments, from fossil findings to fungi observations, or even humanities and social sciences. While AI has already found its way into species identification, on SPOTTERON we are always aiming for adaptability and flexibility in the solutions developed for Citizen Science Apps. With the new SPOTTERON AI, we have created a system to integrate artificial intelligence that can be custom-trained for a large variety of application domains.
Utilizing established machine learning methods, we can train customized AI models to provide image analysis and object/species identification for Citizen Science Apps. Our models run on state-of-the-art server deployments, featuring highly parallelized processing and dynamic scaleability within the SPOTTERON AI system. This allows computational power to be added on-demand in situations of high usage, for example during dissemination campaigns and other phases of increased user submissions. Advanced optimizations reduce resource usage and increase processing speed on our model servers. With this extended AI integration, we provide a stable, performant, efficient and scalable AI solution for user-driven Citizen Science Apps.
With the upgraded SPOTTERON AI system in place, we can now provide custom AI-based image analysis for all kinds of Citizen Science projects, from model training to runtime. The integration of AI in apps for Citizen Science and Citizen Action brings huge potential for user participation and scientists. Especially in more complex topics like the IPM-Popillia App, in which it is crucial to identify an invasive controlled pest species correctly for reporting to authorities, the integration of AI species identification can improve data quality and user experience. With the SPOTTERON AI, we can bring those qualities to national and international Citizen Science projects and to consortiums in Horizon Europe projects.
We are happy to get in contact with your Citizen Science or other interactive projects and develop innovative AI solutions together with your team. SPOTTERON is also available to join as a project partner in Horizon Europe consortiums, or other science funding programmes. Drop us an email and let's schedule a call!
By accepting you will be accessing a service provided by a third-party external to https://www.spotteron.net/