The Institute of Intelligent Robotics at Novosibirsk State University (NSU) has introduced a trio of free online tools that run on the iOk artificial intelligence platform for automatic analysis and identification of images captured with microscopes. This suite was created to relieve researchers from repetitive image-processing chores, enabling them to focus more on interpretation and discovery. The information was shared by the Ministry of Education and Science of the Russian Federation through its press service, and subsequently reported by socialbites.ca.
Researchers often face the need to characterize microscopic images, such as calculating average object size or counting items within an image. In the past, these tasks required manual measurements and time-consuming steps. Andrey Matveev, who leads the Deep Machine Learning in Physical Methods Laboratory at NSU’s Institute of Intelligent Robotics, explained that the new platform aims to streamline these operations, reducing the manual workload and accelerating the pace of analysis while preserving accuracy. The shift from manual to automated processing represents a significant efficiency gain for laboratories that work with a large volume of samples or live-time imaging data.
The iOk platform consists of three AI-powered online services designed for image analysis. These services are named No Code ML, DLgram, and ParticlesNN. No Code ML is implemented as a Telegram bot, offering quick access for researchers who prefer messaging interfaces. DLgram operates as a web-based service, providing a broader, browser-friendly workspace for data processing. ParticlesNN is another web-based tool focused on nuanced object recognition and measurement. All three programs share a common foundation: a Cascade Mask-RCNN neural network that has been trained on a diverse dataset of around 5,000 objects to support robust identification across various imaging conditions.
These tools are built to handle a wide range of image sources. They can analyze images and video recordings produced by electron microscopes as well as standard digital cameras, including those in smartphones. The platform is capable of recognizing multiple object types, such as nanoparticles, microorganisms, cells, and more, enabling researchers to extract meaningful metrics and observations from their visual data with minimal setup.
Matveev notes a key distinction of NSU’s services: they do not demand exceptionally high-resolution images to function effectively. The iOk suite has been developed with resilience in mind, including the ability to identify and suppress noise and lighting variations that could otherwise be misinterpreted as separate objects by some AI systems. This capability improves the reliability of reports and reduces false positives in automated analyses, which is especially important when downstream decisions hinge on precise quantification.
To streamline usage, the developers have integrated the three capabilities into a single, cohesive platform called iOk, enhancing user convenience and workflow efficiency. The growing user base now exceeds five hundred domain experts, highlighting the platform’s practical value across laboratories and research settings where rapid, repeatable image analysis is essential.
Beyond the immediate benefits for microscopy workflows, the NSU efforts contribute to the broader conversation about AI-assisted research. By lowering technical barriers and enabling more researchers to leverage advanced image-analysis techniques, this work supports faster hypothesis testing, improved data quality, and the potential for new insights in materials science, biology, and related fields. The integration of these tools into routine laboratory practice represents a meaningful step toward more automated, data-driven discovery in scientific research.
In related advances, artificial intelligence has already played a role in identifying new therapeutic candidates for anemia, illustrating how AI-enabled analysis can accelerate biomedical research and generate practical outcomes that may improve patient care. The NSU iOk platform exemplifies how institutions can translate cutting-edge machine learning methods into accessible applications that empower researchers to extract reliable information from complex visual data, with broad applicability across disciplines and imaging modalities.