AI enhances security screening – Master’s thesis improves See Through Solutions’ camera technology

See Through Solutions aims for global markets with a solution that modernizes and speeds up security screening. Its terahertz-based thermal camera can accurately and discreetly detect objects hidden under clothing. The solution removes the need for manual inspections, helping large numbers of people move more quickly through locations such as border crossings, transport hubs and large public events.

In the company’s technology, the detection of threat objects, such as weapons, plastic explosives and drugs, is automated using artificial intelligence. However, training suitable AI models for this purpose is challenging: collecting sufficiently large and diverse training datasets from terahertz images is labor-intensive and cannot cover every real-world scenario.

Aalto University computer science student Syed Ashraf developed an AI solution for See Through Solutions in his master’s thesis, improving the detection of threats hidden under clothing.

“In my thesis, I examined a key challenge in terahertz imaging: the lack of high-quality and diverse training data,” Ashraf explains.

“In the project, synthetic data was generated and tested to see how it affects the model’s ability to make accurate predictions,” says Perttu Kajatkari, Software Manager at See Through Solutions, who served as the thesis advisor.

The thesis was funded by the Finnish AI Region (FAIR EDIH) initiative, which supports the adoption of AI in SMEs. Aalto University is one of the ten partners in the FAIR network.

“FAIR brings together augmented reality, artificial intelligence and their industrial applications—exactly the areas our technology builds on. In a way, our camera functions as a ‘sixth sense’, allowing us to see things that humans normally cannot,” says Aarne Oja, CEO of See Through Solutions.

Towards new breakthroughs

The goal of Ashraf’s thesis was to improve the detection performance of the object recognition model used in See Through Solutions’ camera technology.

“We decided to use self-generated synthetic data alongside the original training data when training the model. Our tests showed that this significantly improves the model’s performance,” Ashraf says.

The company is highly satisfied with the results. “The trained neural network developed in the thesis has already been integrated into our camera technology, where it enhances the detection of threats hidden under clothing in security screenings,” Kajatkari notes.

For Ashraf, completing his master’s thesis in a company environment was a valuable experience: “It was motivating to solve a real challenge presented by the company. I learned to report on my work and progress through regular discussions with my supervisor, and I was also given responsibility and independence in my work.”

Ashraf is particularly inspired by AI’s potential to improve people’s lives, for example through disease prediction, improved public safety and the prevention of infrastructure failures.

“It’s also fascinating how the field is constantly evolving and new breakthroughs are emerging. There is always more to learn, and you can never really say that you know everything.”

FAIR expands services for companies

The Finnish AI Region is one of the key actors in Finland’s AI ecosystem. Its activities and services expanded further with the launch of its second operational period at the beginning of 2026.

“We strongly encourage other companies to join FAIR’s activities. In addition to thesis funding, it offers many valuable services. In the future, we plan to make use of its consulting services related to GDPR compliance,” Oja says.

Lauri Savioja, Professor at Aalto University and supervisor of Ashraf’s thesis, highlights the importance of FAIR in enabling research that is relevant to companies.

“For universities, it is a significant competitive advantage that master’s theses address real industrial challenges. At the same time, students gain valuable experience in industry collaboration, which clearly improves their employment opportunities after graduation,” Savioja says.

This news article was originally published on the Aalto University website on 26.3.2026