Video-to-Video AI Takes Center Stage: Jaw-Dropping Capabilities and Growing Adoption
The rapid pace of innovation over the past year in the AI space has been astonishing, with video-to-video generative AI emerging as the next frontier. Lightricks, the developer behind the Videoleap app, has showcased remarkable capabilities that push the boundaries of the mobile app industry.
Video-to-video generative AI technology allows users to transform the style of their recorded or uploaded videos, creating something entirely new. By using prompts such as 80s punk or Japanese manga, users can completely change the look and feel of their videos. Additionally, Videoleap enables video transformation based on uploaded images or style presets like Abandoned for a post-apocalyptic aesthetic or Fire to add flickering flames in an anime style.
Text-to-image generators were once ridiculed for their hallucination issues, but significant progress has been made in their stability. A recent study conducted by Lightricks in collaboration with YouGov found that 92% of creators are already well-acquainted with generative AI, and 62% are already utilizing it in their work.
However, concerns about the widespread adoption of AI remain, particularly in relation to the proliferation of manipulative deep fakes. While video-to-video AI cannot convincingly transform everyday users into performing magic tricks like Tom Cruise, for example, the potential for illicit use cases is a valid worry. Lightricks co-founder and CEO Zeev Farbman emphasizes the necessity of critical thinking when consuming visual content and believes that making the technology widely available is vital in addressing this issue.
Video-to-video generators are still in an emergent phase, with ongoing refinement and improvement of underlying algorithms. Currently, the main limitation lies in the server-side resources required to generate each frame of a video, resulting in longer wait times. However, industry progress is evident, with various players in the market offering competing solutions. Lightricks recently introduced AI Effects on their Videoleap app, while Picsart offers a limited video-to-video tool. Both apps have a time limit on video length, but as competition drives innovation, it is only a matter of time before the limitations are extended to accommodate longer content.
Lightricks has pushed the boundaries further with Videoleap, incorporating on-device editing of AI-generated images, which significantly reduces rendering time. They have also introduced a mobile AI Infinite Zoom feature, allowing users to expand their videos using outpainting techniques.
As demand for video-to-video AI tools continues to grow, it will be intriguing to see which company emerges as the market leader. Each player has a slightly different focus, which will attract specific user groups. For instance, Runway aims to capture the premium end of the market, particularly among filmmakers, while Picsart targets creative users with their budget-friendly tool suite.
With a broad user base of 30 million individuals utilizing their products every month, Lightricks is well-positioned to capture a significant share of the demand for video-to-video AI tools. As the video-to-video generative AI space gains momentum, users can experiment with the different features offered by various platforms. Soon, these tools could become commonplace for both businesses and consumers alike.
