Video Content Streaming via Captioned Still Images

Solution Overview

Problem

Users are reluctant to replay video content due to short data capacity, high initial loading times, difficulty in using earphones, and uncontrollable full-screen playback, leading to decreased video-based advertisement exposure and increased user reaction costs.

Innovation Solution

A system that extracts still images from video content, generates corresponding audio scripts, adds captions, and provides caption-added still images in real-time streaming, allowing for efficient content consumption and reduced data usage, while enabling video replay and sharing functions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If video content is provided to users, then content consumption quality is improved, but data usage increases and loading time increases

Engineering Contradiction:

Improvecontent consumption qualityVSAvoiddata usage

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts key visual information from video content by generating still images with captions that represent essential scene elements. This extraction process separates the core informational content from the full video stream, allowing users to consume content through image-based representations that use significantly less data while maintaining comprehension of the original video material.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system creates simplified copies of video content in the form of still images with overlaid captions. These image-based copies capture the essential information of video scenes without requiring users to download or stream the complete video file, thereby reducing data consumption while preserving content delivery.

Inventive Principle:

Principle #26Copying

2Reliability

If video content is provided to users, then content consumption quality is improved, but initial loading time increases

Engineering Contradiction:

Improvecontent consumption qualityVSAvoidinitial loading time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts key visual information from video content by generating still images with captions that represent essential scene elements. This extraction process separates the core informational content from the full video stream, allowing users to consume content through image-based representations that use significantly less data while maintaining comprehension of the original video material.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system performs preliminary processing by pre-generating still images and captions from video content before user consumption. These pre-processed image-based representations are ready for immediate delivery, eliminating the need for users to wait for lengthy video loading while the backend continues to process or update content asynchronously.

Inventive Principle:

Principle #10Preliminary action

3Reliability

If video content is provided to users, then content consumption quality is improved, but user control capability deteriorates

Engineering Contradiction:

Improvecontent consumption qualityVSAvoiduser control capability

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The patent segments video content into discrete still image frames with associated captions, allowing users to navigate through content in controlled increments. This segmentation enables users to pause, review, and control their consumption pace without being locked into continuous video playback, significantly improving ease of operation while maintaining content quality.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9906820B2Method and system for providing video content based on image

Publication Date: 2018.02.27 PIAMOND CORP

AI summary

Disclosed is a method for providing a content, the method including extracting at least one still image from video included in the content, extracting audio, which corresponds to the still image, and generating a script corresponding to the audio, adding a caption to the still image based on the generated script, and providing the content in response to a request of consumption to the content and providing the caption-added still image for the video that is streaming in real time.