Students Developed Tools for Content Modernisation in Web
Students of Moscow Poly and National University of Science and Technology MISiS became winners of SberZvuk Tech Days hackathon. They provided the best solution for development of algorithm for filtration of USG, user generated content.
In a short time, students provided a stable code for voice recognition of various people on audio recordings, as well as human faces recognition on videos. An ability to censor the chosen parts of the content is one the key features, as is variability in sound processing.
Tools can be applied for simplification of searching info on actors and directors on streaming services, as well as during development of tag list for video and audio podcasts. This will allow improving navigation in preferred content.
In the long run it is possible to integrate this algorithm in SberZvuk system, developed by hackathon organisers, as well as increase data base of public figures to increase app’s accuracy of search.
Moscow Poly student Yana Richlitzkaya participated in the development