GCC: AI 기반 음악 장르 조합 검색 엔진

Existing music search and recommendation rely on vague tags such as "Genre=Rock" and "Mood=Exciting," making it difficult to reflect the detailed characteristics users want. For example, when a creator wants to find a video BGM that is "a bit fast, punchy, and without vocals," there are limitations to finding music that fits the purpose with text tags.

In addition, genre and mood tags are subjective and lack consistency, resulting in large deviations in search results, and it takes a lot of time to select the exact song from a wide range of music catalogs. In summary, inefficient text-based music search makes it difficult for both content creators and listeners to explore the music they want.

GCC analyzes the actual audio signal of a sound source with AI, allowing music to be decomposed and combined into detailed sound element units beyond genre. For example, it is possible to search for songs with a specific set of conditions such as "fast tempo + strong beat + lots of electronic sounds + no female vocals + high sense of space." The UI visualizes 9 main components and 6 auxiliary elements with graphs and buttons to support intuitive combination search.

This enables exploration based on multi-dimensional music characteristics that were previously impossible, allowing users to define their own custom genres to find music. In addition, the AI algorithm implemented with 9 proprietary patented technologies (based on RSL) has excellent search speed and accuracy, capable of recommending desired songs within 1 second, realizing real-time filtering even in vast music DBs.

Content creators (B2C/B2B) in video, game, and metaverseandMusic producers and labels (B2B)are the main users. When YouTubers or advertisement creators look for background music, they can easily find songs with the atmosphere they want by defining them with objective characteristics using the GCC engine.

In addition, KEISER has launched its own platform, JIGPU, allowing creators to use and trade music found through GCC search without worrying about copyright issues. In other words, music library providers or OTT/music streaming services can also license this search engine and incorporate it into recommendation algorithms. In fact, technology integration contracts have also been signed with other AI solutions (such as mental care apps).

The GCC engine can be applied to the entire global music catalog without language barriers, so it can be expanded to streaming platforms and music databases around the world. In addition, since it supports secure music transactions in connection with blockchain-based copyright management, expansion is expected into music IP businesses such as music distribution and investment platforms.

There are no special restrictions in terms of regulations, but securing a large-scale music data license and the participation of musicians are essential for successful establishment. Currently, it has secured about 1,000 songs from about 500 musicians and is preparing contracts with overseas artists in the United States and Vietnam. Based on this, it plans to expand global services starting with CES 2026.

It was awarded the Innovation Award with the evaluation of "presenting a new paradigm for music search." The CES judges highly praised GCC for decoding the DNA of music with AI, and especially praised it as a practical tool for content creators.

Regarding technical completeness, there were positive opinions that "the algorithm reliability is high due to 15 years of research accumulation," and it was also mentioned that entry barriers were established by holding domestic and foreign patents. However, in terms of the market, there is a point that "securing partnerships with music platforms is a challenge." Although the innovation is outstanding, it is uncertain how much the GCC engine will be adopted as an industry standard, as large streaming companies are upgrading their own recommendation systems.