Russian Teen Builds AI App That Detects Emotions from Photos

Backed by a national tech incubator, this high school project could shape the future of education and cognitive research.

In a country-wide push for tech innovation, a Russian high school student is breaking new ground in AI-powered emotion recognition. Victoria Dikova, a junior at Ulyanovsk Lyceum No. 20, is developing FaceScope, a mobile application that reads human emotions from a photograph—an idea that could transform fields like remote education and psychological research.

The project is supported by Stellar, a resident startup of the Sirius Science and Technology Center—Russia’s answer to Silicon Valley. With FaceScope, users can upload a photo and receive a real-time emotional analysis, accompanied by a short explanation. So far, the AI handles basic emotional states—think joy or sadness—but Dikova says more nuanced signals like sarcasm still require refinement.

“We’re training the model to understand not just expressions, but intent,” says Dikova. The development team chose Lua, a lightweight scripting language known for its simplicity and resemblance to Python, to code the first working prototype.

FaceScope is part of a broader push to apply machine learning in real-time learning environments. Teachers using remote platforms, for example, could gain insights into how students are reacting during lessons—whether they’re engaged, confused, or anxious. Cognitive researchers could also use the app for emotional profiling in studies without needing invasive equipment or extensive training.

The finished prototype will be demoed later this year at the Big Challenges Festival hosted by Sirius, where visitors can put it to the test. It’s a small project by global standards—but it’s powered by the big ambition of a 17-year-old aiming to make AI more empathetic.