Video system with scene-based object insertion feature

A computing system configured for performing a set of acts comprising:

• obtaining video depicting an area across multiple frames of the video, wherein the area is part of a scene of the video, and wherein the area is suitable for having an object inserted therein;
• detecting the area within the obtained video and determining area characteristic data associated with the detected area;
• determining scene attribute data associated with the scene, wherein determining scene attribute data associated with the scene comprises using video data representing the video to determine the scene attribute data;
• using at least the determined area characteristic data and the determined scene attribute data as a basis to select an object to be inserted into the detected area, from among a set of multiple candidate objects;
• obtaining description data associated with the selected object to be inserted into the detected area, wherein the obtained description data provides an indication of the importance of the object's scale;
• inserting the selected object into the detected area to generate video that is a modified version of the obtained video, wherein inserting the selected object comprises:
• (i) obtaining a three-dimensional model of the selected object;
• (ii) scaling the three-dimensional model of the selected object and using the provided indication of the importance of the object's scale as a basis to select an extent of resources to utilize in connection with scaling the three-dimensional model of the object being inserted into the detected area;
• (iii) using the obtained three-dimensional model of the selected object and the determined area characteristic data, together with a time-based transform model, to generate a time-based two-dimensional projection of the selected object;
• (iv) determining area position data associated with the detected area;
• and (v) at a position indicated by the determined area position data, inserting into the detected area the corresponding time-based two-dimensional projection of the selected object;
• and
• outputting for presentation the generated video.

A non-transitory computer-readable medium having stored thereon program instructions that upon execution by a computing system, cause performance of a set of acts comprising:

• obtaining video depicting an area across multiple frames of the video, wherein the area is part of a scene of the video, and wherein the area is suitable for having an object inserted therein;
• detecting the area within the obtained video and determining area characteristic data associated with the detected area;
• determining scene attribute data associated with the scene, wherein determining scene attribute data associated with the scene comprises using video data representing the video to determine the scene attribute data;
• using at least the determined area characteristic data and the determined scene attribute data as a basis to select an object to be inserted into the detected area, from among a set of multiple candidate objects;
• obtaining description data associated with the selected object to be inserted into the detected area, wherein the obtained description data provides an indication of the importance of the object's scale;
• inserting the selected object into the detected area to generate video that is a modified version of the obtained video, wherein inserting the selected object comprises:
• (i) obtaining a three-dimensional model of the selected object;
• (ii) scaling the three-dimensional model of the selected object and using the provided indication of the importance of the object's scale as a basis to select an extent of resources to utilize in connection with scaling the three-dimensional model of the object being inserted into the detected area;
• (iii) using the obtained three-dimensional model of the selected object and the determined area characteristic data, together with a time-based transform model, to generate a time-based two-dimensional projection of the selected object;
• (iv) determining area position data associated with the detected area;
• and (v) at a position indicated by the determined area position data, inserting into the detected area the corresponding time-based two-dimensional projection of the selected object;
• and
• outputting for presentation the generated video.