Client-Matched Video Stream Selection for Personalized Live Streaming

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video stream processing methods fail to provide personalized video stream delivery to different clients due to varying user preferences, device capabilities, and viewing requirements, leading to a lack of tailored live-streaming experiences.

Innovation Solution

A method and apparatus that analyze multiple video streams for a live-streaming event, determine feature information including shooting position and content, match this information with client-specific preferences and configurations, and select and push the most suitable stream based on calculated matching degrees.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple video streams are provided for the same live-streaming event, then the live-streaming content is enriched and viewing experience is enhanced, but it becomes difficult to meet diverse individual preferences of different users

Engineering Contradiction:
Improvelive-streaming content enrichmentVSAvoidvideo stream selection complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the video stream selection process into distinct components: feature extraction from video streams, feature extraction from client preferences, matching degree calculation, and final selection. This segmentation transforms the complex overall problem into manageable sub-tasks that can be processed systematically

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameters of video streams by extracting meaningful features (shooting position, content characteristics) and transforms client preferences into comparable feature representations. This parameter transformation enables quantitative comparison and matching between multiple video streams and client preferences

Inventive Principle:
Principle #35Parameter changes

2Ease of operation

If video streams are selected based on user preferences, then personalized viewing experience is achieved, but the processing complexity and time consumption increase

Engineering Contradiction:
Improvepersonalized viewing experienceVSAvoidvideo stream processing time
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent performs preliminary action by pre-extracting features from video streams before actual client requests arrive. Video stream features such as shooting position and content characteristics are analyzed in advance, so that when a client requests a stream, the matching process can proceed quickly using pre-computed features

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates feature representations (copies) of both video streams and client preferences in a common feature space. Instead of comparing raw video data with client preferences, the system works with compressed feature vectors that capture essential characteristics, significantly reducing processing complexity and time

Inventive Principle:
Principle #26Copying

3Measurement precision

If feature matching is performed between video streams and clients, then accurate stream selection is achieved, but the computational load and system complexity increase

Engineering Contradiction:
Improvestream-client matching accuracyVSAvoidmatching system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts essential features from both video streams (shooting position, content type) and client preferences (viewing habits, device characteristics), separating the critical matching criteria from irrelevant data. This extraction focuses computational resources on the most important features that actually determine matching quality

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces feature vectors as an intermediary representation between raw video streams and client preferences. These feature vectors serve as a common language that enables comparison, acting as a mediator that transforms diverse inputs into a unified format suitable for matching degree calculation

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4149115B1Video stream processing method and apparatus, and computer device and medium
Publication Date: 2025.09.10 BAIDU ONLINE NETWORK TECH (BEIJIBG) CO LTD
  • EP4149115B1 patent drawingFigure 1
  • EP4149115B1 patent drawingFigure 2
  • EP4149115B1 patent drawingFigure 3A

AI summary

The present disclosure provides a method of processing a video stream, which relates to a field of a cloud platform. The method includes: acquiring a plurality of video streams for a live-streaming event; determining respective feature information of the plurality of video streams; determining a feature information of a client in response to a video request message from the client; matching the respective feature information of the plurality of video streams with the feature information of the client respectively, so as to obtain respective matching degrees of the plurality of video streams with respect to the client; and selecting one video stream from the plurality of video streams to obtain a selected video stream according to the respective matching degrees of the plurality of video streams with respect to the client, and pushing the selected video stream to the client. The present disclosure further provides an apparatus of processing a video stream, a computer device, and a medium.