Media editing with client-side edits and server-side edits

A system, comprising:

• memory;
• one or more processors coupled to the memory and configured to:
• receive on a client a specification of an edit to a portion of a video;
• provide on the client a real-time preview of the video with the edit applied dynamically to the portion of the video prior to encoding the video, wherein dynamically applying the edit comprises:
• applying the edit to a first individual video frame of the plurality of individual video frames corresponding to the portion of the video based on the received specification of the edit;
• and
• displaying the first individual video frame of the plurality of individual video frames before applying the edit to a second individual video frame of the plurality of individual video frames corresponding to the portion of the video based on the received specification of the edit;
• provide input data to a machine learning model, wherein the machine learning model is used to make a determination of applying the edit to the portion of the video on the client or a server of a cloud service;
• in response to the determination of applying the edit to the portion of the video on the server of the cloud service, provide to the server of the cloud service at least the specification of the edit to the portion of the video and the portion of the video without the edit applied;
• generate, based at least in part on the determination, an edited video comprising the portion with the edit;
• in response to the determination of applying the edit to the portion of the video on the server of the cloud service, applying the edit to the portion of the video on the client and providing to the server of the cloud service the portion of the video with the edit applied;
• and
• in response to providing to the server the portion of the video with the edit applied, cause the video to be deleted.

A computer program product embodied in a non-transitory computer readable medium and comprising computer instructions for:

• receiving on a client a specification of an edit to a portion of a video;
• providing on the client a real-time preview of the video with the edit applied dynamically to the portion of the video prior to encoding the video, wherein dynamically applying the edit comprises:
• applying the edit to a first individual video frame of the plurality of individual video frames corresponding to the portion of the video based on the received specification of the edit;
• and
• displaying the first individual video frame of the plurality of individual video frames before applying the edit to a second individual video frame of the plurality of individual video frames corresponding to the portion of the video based on the received specification of the edit;
• providing input data to a machine learning model, wherein the machine learning model is used to make a determination of applying the edit to the portion of the video on the client or a server of a cloud service;
• in response to the determination of applying the edit to the portion of the video on the server of the cloud service, providing to the server of the cloud service at least the specification of the edit to the portion of the video and the portion of the video without the edit applied;
• generating, based at least in part on the determination, an edited video comprising the portion with the edit;
• in response to the determination of applying the edit to the portion of the video on the client, applying the edit to the portion of the video on the client and providing to the server of the cloud service the portion of the video with the edit applied;
• and
• in response to providing to the server the portion of the video with the edit applied, causing the video to be deleted.