Naver Cloud announced on the 21st that it introduced the video analysis service 'MAIU' at the International Broadcasting, Media, Sound, and Lighting Exhibition (KOBA 2025) held at COEX in Seoul, which analyzes information in videos using artificial intelligence (AI) to automatically extract metadata and support searches.
The service is currently in closed beta service (CBT) for domestic broadcasters and partners and is scheduled for formal launch next month.
This service automatically analyzes various elements in videos, such as people, actions, backgrounds, and voices, to generate metadata, helping users quickly find desired scenes through text searches. For example, scenes like 'the scene with a wine glass' and 'the scene where the male protagonist runs' can be quickly located through text searches alone.
Company officials explained that this feature can significantly reduce the time required for metadata extraction and rough editing.
According to Naver Cloud, if a single episode of a variety show is filmed with about 3,600 minutes of video, it previously took the editing team about 32 hours (3 to 4 days) for analysis and rough editing. However, by utilizing MAIU, this process can be shortened to within 2 hours, which means editing time can be reduced by more than 15 times, according to company representatives.