Students of the University of Transport Technology (UTT) Participate in the International Competition OmniSub 2026
Within the framework of the International Science Festival “CIFR-2026” organized by the Moscow Technical University of Communications and Informatics (MTUCI), the delegation of the University of Transport Technology (UTT), led by Mr. Phạm Đức Anh and Mr. Trần Văn Tâm, accompanied and supported two student teams from the Faculty of Information Technology to participate in the international competition OmniSub 2026 – Visual Subtitles Challenge. The competition was organized by the Artificial Intelligence Research Institute (AIRI)—one of the leading AI research centers in the Russian Federation—in collaboration with MTUCI.
OmniSub 2026 is a highly competitive international hackathon, bringing together undergraduate, graduate, and doctoral students from countries within the CIS, BRICS, Cuba, and Vietnam. Beyond its academic significance, the competition also features an attractive prize structure, with the First Prize valued at 200,000 RUB, the Second Prize at 100,000 RUB, and the Third Prize at 50,000 RUB. This prize scale reflects a highly professional organization and creates a competitive environment where teams must continuously improve their models to enhance rankings. In addition, the evaluation system is based on strict criteria such as subtitle accuracy (WER, CER), recognition capability under complex conditions, and processing performance, requiring teams to deeply optimize algorithms, data, and generalization capacity.
Information about OmniSub 2026
A distinctive aspect of the competition lies in the challenge of developing an automatic subtitle generation system for videos without spoken audio. To address this problem, teams were required to integrate multiple domains such as facial recognition, lip-reading analysis, and natural language processing to infer content. This represents a multimodal artificial intelligence problem, where visual and linguistic data are closely integrated, reflecting advanced research trends with high practical application potential.
The two UTT student teams participated in the online competition from March 16 to March 19, 2026, competing directly with strong teams from technical universities in the region. In their first participation in a highly demanding international competition, the students had to quickly grasp the problem requirements while simultaneously building, training, and optimizing models within a limited timeframe. Time pressure, high technical requirements, and continuous competition on the leaderboard created a comprehensive challenge in both professional expertise and teamwork skills.
As a result, team VN.UTT1, under the supervision of Mr. Phạm Đức Anh and Mr. Trần Văn Tâm, including students Hạ Nguyễn Mạnh Hùng, Trần Minh Khoa (Class 73DCTT24), Đới Sỹ Nam, and Ninh Thị Thu Huyền (Class 74DCTT26), achieved an outstanding ranking of 6th out of 23 participating teams. This achievement exceeded expectations for a team participating for the first time in a highly specialized international competition such as OmniSub.
Results of OmniSub 2026
Notably, achieving a Top 6 position not only reflects the final outcome but also demonstrates the team’s ability to maintain consistent performance throughout the competition, as teams continuously updated models and improved scores. This achievement shows that UTT students are fully capable of competing with teams from technologically advanced countries, while also affirming the effectiveness of the University’s practice-oriented and research-driven training approach.
Beyond rankings, the competition provided significant practical value for students. Through participation, teams gained hands-on experience in developing a complete AI system—from data processing and model development to evaluation and optimization. Students also strengthened their ability to work under pressure, enhancing critical thinking, problem-solving, and teamwork skills—key competencies in modern technology fields.
In terms of application, the solutions developed in the competition hold great potential in supporting hearing-impaired individuals in accessing digital content, automating subtitle generation for videos, and improving information accessibility in the digital society. This also demonstrates how AI research problems are increasingly connected to real-world challenges and deliver meaningful social value.
The achievement at OmniSub 2026 is not only a memorable milestone for participating students but also contributes to affirming the position of the University of Transport Technology in artificial intelligence training and research. In the coming time, the experience gained from the competition will serve as an important foundation for students to participate in larger academic arenas, develop advanced research, and contribute to the advancement of technology both in Vietnam and internationally.
University of Transport Technology
