AI-focused startup, Memories.ai, secures $8 million in seed funding for the development of visual solutions for artificial intelligence.
Revolutionary AI Technology Enables Persistent Visual Memory
In a groundbreaking development, Memories.ai, an AI research lab, has unveiled the world's first Large Visual Memory Model (LVMM). This innovative technology enables AI to ingest, index, and recall vast volumes of video data over extended timeframes, transforming the way industries handle and analyse visual information [1][2][3][4].
The LVMM addresses a major limitation of previous AI systems, which could only retain video context for very short periods (15-60 minutes) [2]. With this technology, AI can now capture, store, and organize visual data over time, enabling them to recognise patterns and compare new footage to past events [1][2][3][4].
The current and potential applications of the LVMM span several industries. In security and safety, AI equipped with Memories.ai's technology can search months of surveillance footage in seconds, detect subtle anomalies, and identify recurring patterns to enhance threat detection and safety protocols [1][2][3][4]. In media and entertainment, the LVMM can locate specific scenes, props, or characters in decades of archived content, drastically reducing editing, indexing, and licensing time for content creators and producers [1][2][4].
In marketing and analytics, the LVMM can analyse sentiment and trending topics across millions of social video clips, track brand visibility, and consumer engagement [1][3][4]. For robotics, the LVMM enables robots to have human-like long-term visual memory to understand and recall environmental and operational contexts over time [1][3]. In consumer devices, the LVMM can be integrated with mobile devices (Samsung collaboration) to allow users to search and recall personal video memories seamlessly [1][4].
Memories.ai's platform is accessible via both a developer API and a chatbot app, enabling diverse clients—from security firms to marketers and device manufacturers—to upload or integrate video libraries for instant querying and actionable insight generation [1][3][4]. The company’s recent $8 million seed funding led by investors including Samsung Next underlines strong confidence in LVMM’s transformative impact across sectors [3][4].
In summary, the LVMM enhances AI’s ability to "see, understand, and recall" visual information persistently and contextually, unlocking new possibilities in video-intensive industries by turning vast archives into searchable, meaningful, and actionable memories [1][2][3][4]. As the technology matures, ongoing expansions of use cases and partnerships are anticipated.
References:
[1] Memories.ai. (n.d.). Retrieved from https://memories.ai/
[2] VentureBeat. (2023, March 23). Memories.ai raises $8M to unlock AI's long-term visual memory. Retrieved from https://venturebeat.com/2023/03/23/memories-ai-raises-8m-to-unlock-ais-long-term-visual-memory/
[3] TechCrunch. (2023, March 23). Memories.ai raises $8M to help AI remember what it sees. Retrieved from https://techcrunch.com/2023/03/23/memories-ai-raises-8m-to-help-ai-remember-what-it-sees/
[4] The Information. (2023, March 23). Memories.ai raises $8 million for its AI that can remember what it sees. Retrieved from https://www.theinformation.com/articles/memories-ai-raises-8-million-for-its-ai-that-can-remember-what-it-sees
Artificial intelligence, powered by the Large Visual Memory Model (LVMM), can now remember and understand visual information persistently, allowing for swift searching and analysis of video archives in various industries. This revolutionary technology demonstrates the potential of artificial-intelligence to transform not only how industries handle and analyze video data but also the ways robots, consumer devices, and more can leverage long-term visual memory.
