Video Frame Selection for Targeted Content Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Targeted advertising based on video content has been ineffective, particularly when videos are montaged or divided, and relies heavily on computationally intensive object recognition.
Innovation Solution
A system that allows target content providers to select specific video frames for targeting, enabling precise content delivery irrespective of playback timing, by detecting and communicating relevant content to users' devices for presentation as overlays or on secondary devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If object recognition is used to target advertisements in video content, then advertisement targeting accuracy is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent extracts specific visual elements (objects, scenes, actions) from video content and represents them as discrete tags or metadata. Instead of performing continuous object recognition on entire video frames, the system identifies and extracts key visual concepts as separate data points that can be efficiently matched against advertisement criteria, reducing computational complexity while maintaining targeting accuracy.
Solution Approach 2:
The patent creates simplified representations (copies) of video content in the form of metadata tags, scene descriptions, and object labels. These copied representations serve as proxies for the actual video content, allowing advertisement targeting to be performed on the lightweight metadata rather than the computationally intensive original video data, thus reducing processing requirements while preserving targeting capability.
2Ease of operation
If time markers are used to target advertisements in video streams, then advertisement placement is simplified, but effectiveness decreases when videos are montaged or divided into segments
Solution Approach 1:
The patent creates a universal tagging system that works across different video formats and delivery methods. By tagging visual content with descriptive metadata that is independent of video structure, the system enables consistent advertisement targeting whether the video is delivered as a complete stream, divided into segments, or reassembled through montage. The same tag-based approach functions universally across all these scenarios.
Solution Approach 2:
The patent performs preliminary tagging of video content during ingestion or preprocessing, creating a metadata layer that describes visual elements before the video is delivered or manipulated. This preliminary action ensures that advertisement targeting information is already prepared and attached to the appropriate visual content, so subsequent montage, segmentation, or reassembly operations do not require re-processing or lose targeting accuracy.
3Measurement precision
If all video content is pre-processed for advertisement targeting, then targeting precision is improved, but processing time and resource consumption increase
Solution Approach 1:
The patent applies partial processing by selectively tagging only the most visually significant or advertisement-relevant portions of video content. Instead of comprehensively analyzing every frame and visual element, the system identifies and tags key objects, scenes, or actions that are most likely to be relevant for advertisement targeting. This partial action approach achieves sufficient targeting precision while dramatically reducing processing time and resource requirements compared to exhaustive analysis of all video content.
Data Source
AI summary
Systems, methods, and computer-readable storage media are provided for providing target content, such as advertisements, based on one or more selected video frames. A set of video frames and target content is received. The target content is to be presented upon detection of a playback of the set of video frames. The playback of the set of video frames is detected. In response to the detection of the playback of the set of video frames, the target content is communicated for presentation.