Case Study: SBS Optimizes Special Effects Archive with TwelveLabs

SBS streamlines scene-level search with TwelveLabs across 3 phases

SBS, a leading South Korean television and radio broadcaster, faced significant challenges in managing its vast video archive. Locating specific special effects or scenes was a difficult and time-consuming process, compounded by a lack of scene-level search capabilities and unaccounted-for media uploaded by individual team members. To address this, SBS partnered with vendor Twelve Labs to implement its multimodal AI technology.

Twelve Labs implemented a multi-phase solution using its Marengo and Pegasus models to enable semantic, scene-level search across SBS's archives. This allowed for the automated identification of VFX elements and key moments, drastically reducing the time spent on content retrieval. The solution provided SBS with immediate access to searchable video content, enabling the efficient reuse and repurposing of archived assets to meet domestic and international demand.

SBS