Video Key Role Detection for Poster Generation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for creating promotional materials for videos, such as posters, are inefficient and costly due to reliance on metadata, which is often lacking in IPTV and VOD systems, and require manual skills, making them impractical for large volumes of videos with limited metadata.

Innovation Solution

Automatically identifying key roles and their relationships within a video by segmenting it into hierarchical structures, performing face detection and grouping, and constructing a community graph to generate visual summaries like posters without relying on metadata.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If conventional metadata-based methods are used to identify key roles, then the quality of promotional materials can be maintained, but the process becomes time-intensive and expensive requiring manual skills

Engineering Contradiction:
Improvepromotional material creation efficiencyVSAvoidmanual skill requirement
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The system performs automatic video content analysis to identify key roles, characters, and scenes without requiring manual intervention. The automated approach uses computer vision and natural language processing to extract promotional content elements, eliminating the need for manual skills while maintaining productivity

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

Manual graphic design and content analysis processes are replaced with automated computational systems. The patent substitutes human operators with machine learning models that automatically analyze video content, extract key elements, and generate promotional material specifications

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If metadata-rich videos are analyzed using conventional methods, then accurate key role identification is achieved, but the approach is not practical for IPTV and VOD systems with limited metadata

Engineering Contradiction:
Improvekey role identification accuracyVSAvoidmetadata availability requirement
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent extracts key information directly from video content through automated analysis rather than relying on external metadata. By taking out the dependency on metadata and extracting information directly from visual and audio content, the system achieves accurate key role identification for both metadata-rich and metadata-poor videos

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The automated content analysis system is designed to work universally across different video types and metadata availability conditions. The same system can process metadata-rich traditional videos and metadata-poor IPTV/VOD videos, making the solution adaptable to various scenarios without requiring separate approaches

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Manufacturing precision

If manual techniques are used to create promotional posters, then quality posters can be generated, but the process becomes difficult to administer as video volume increases

Engineering Contradiction:
Improvepromotional poster qualityVSAvoidprocess administration complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The system automatically generates promotional poster specifications by analyzing video content and identifying key visual elements, characters, and scenes. This self-service approach eliminates the need for manual graphic design processes, maintaining poster quality while simplifying administration as video volumes scale

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent segments video content into analyzable components (scenes, shots, key frames) and processes them systematically. This segmentation allows automated identification of promotional elements without requiring complex manual review processes, making the system scalable to large video volumes

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS9271035B2Detecting key roles and their relationships from video
Publication Date: 2016.02.23 MICROSOFT TECHNOLOGY LICENSING LLC
  • US9271035B2 patent drawing
  • US9271035B2 patent drawing
  • US9271035B2 patent drawing

AI summary

Tools and techniques for acquiring key roles and their relationships from a video independent of metadata, such as cast lists and scripts, are described herein. These techniques include discovering key roles and their relationships by treating a video (e.g., a movie, television program, music video, and personal video, etc.) as a community. For instance, a video is segmented into a hierarchical structure that includes levels for scenes, shots, and key frames. In some implementations, the techniques include performing face detection and grouping on the detected key frames. In some implementations, the techniques include exploiting the key roles and their correlations in this video to discover a community. The discovered community provides for a wide variety of applications, including the automatic generation of visual summaries or video posters including acquired key roles.