Auto-Reframing And Multi-Cam Functions Of Video Editing Application

Smart overview of the Invention

The application describes a computing device configured to modify videos using a machine learning model. The device utilizes multiple training data sets, each containing an original and a modified video. By cropping frames from the original video, the device generates corresponding frames in the modified video. The trained machine learning model is then applied to new videos to produce modified versions, enhancing video editing efficiency, especially for users with limited experience.

Technical Field

This innovation focuses on automatic video editing, particularly through machine learning models that crop and alter videos. It also involves generating media compositions from multiple camera angles, aiming to simplify video editing on touchscreen devices. This approach addresses the challenges faced by both novice and experienced editors using standard tablet equipment.

Background

Video editing on tablets can be challenging due to the limitations of touchscreen interfaces and the complexity of traditional editing tools. Automated editing and intuitive interfaces can streamline this process, making it more accessible and less error-prone. The application seeks to improve the editing experience by leveraging machine learning to automate cropping and merging tasks, thus reducing the effort required from users.

Functionality

The computing device can display multiple media streams and a merged composition, showing different perspectives of an environment. Users can select streams via an interface to create a merged composition. The device allows selection of audio, video, or combined streams, generating compositions based on user choices. This flexibility enables users to create relevant and engaging video content efficiently.

Advantages

The described system offers significant benefits, such as intelligent cropping and zooming to highlight important video portions, regardless of varying source video aspect ratios. Users can create merged compositions from multiple perspectives, enhancing the viewing experience. The user-friendly interface supports real-time composition generation, making video editing more intuitive and accessible.