SK Telecom (hereinafter referred to as SKT) announced on the 13th that it will invest $3 million in the AI video analysis startup 'Twelve Labs.'
Twelve Labs is a startup developing a multimodal AI model that can understand and search videos, established in San Francisco, USA, in 2021. The company gained attention this year after successfully raising about $50 million in investment from NVIDIA's subsidiary NVentures and New Enterprise Associates (NEA) in June.
Twelve Labs identifies various people and objects in videos and understands their actions through its advanced AI video understanding model. Based on this, it provides a video search solution called 'Marengo' and a solution that can answer queries from videos called 'Pegasus.'
Twelve Labs is gaining global attention, having been selected as one of the Global 100 AI Startups 'AI100' by the global research firm CB Insights for three consecutive years from 2022 to 2024.
SKT and Twelve Labs expect to provide customers with more innovative AI experiences by combining their respective AI capabilities to enhance services such as 'AI agents.'
For example, if requests such as 'Find the video of a baby taking its first steps' or 'Show me videos of football players Son Heung-min and Kim Min-jae passing' or 'Summarize what is happening in this video in text' are made among thousands of videos, it can accurately find the desired scenes and summarize the content of the videos in text.
In addition, both companies decided to work together on developing and enhancing related technologies for applying multimodal AI in security and public safety fields, such as AI monitoring systems.
In the case of AI monitoring systems, unlike before, when a single operator had to monitor numerous CCTV cameras for extended periods, applying Twelve Labs' multimodal AI model allows for quick searching and summarization of major incidents, movements, and individuals that occurred that day.
In the case of CCTV integrated control centers, the demand for high levels of real-time detection and searching efficiency in challenging conditions, such as bad weather or poorly lit streets, is increasing, making opportunities for utilizing multimodal AI seem limitless.
Twelve Labs also decided to join the 'K-AI Alliance' led by SKT to collaborate on various synergy projects with member companies, contributing to the activation of the AI ecosystem. Global AI corporations are investing in and researching multimodal technologies, which can analyze and generate various forms of data, including videos, voice, and images, beyond just text. Global market research firm 'Mordor Intelligence' has projected that the AI video analysis market size will reach $75.35 billion by 2028.
Lee Jae-sung, CEO of Twelve Labs, stated, 'I expect that collaboration with SK Telecom will allow Twelve Labs' video understanding technology to provide innovative video understanding AI solutions to more industrial sites.'
Lee Jae-shin, head of SKT's AI growth strategy division, noted, 'Through collaboration between the two companies, we will strengthen our competitiveness in the multimodal AI field' and added, 'We plan to expand services and business areas where customers can truly feel AI innovation.'