Livestream Video Event Detection for Automated Esports Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional esports platforms face limitations in tracking competitor scores and verifying achievements in race format competitions, as they rely on manual reporting by competitors or human administrators, which restricts the number of participants and accuracy.
Innovation Solution
Employing machine vision technologies trained to recognize events in livestream videos of esports competitions, automatically detecting occurrences of events across multiple video games and updating scores without manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual reporting by competitors or human administrators is used to track scores, then verification of achievements can occur, but the number of participants is restricted and accuracy is compromised
Solution Approach 1:
The patent replaces the manual mechanical system of human administrators watching videos and noting achievements with an automated computer vision system. The system uses machine learning models to automatically detect events in livestream videos, extract relevant information, and update scores without human intervention. This substitution enables the system to handle a much larger number of participants simultaneously while maintaining or improving accuracy through consistent automated detection.
Solution Approach 2:
The system enables self-service by allowing competitors to livestream their gameplay automatically, with the automated system extracting achievement information directly from these streams without requiring competitors to manually report or provide proof. The livestreaming service and automated detection work together to verify achievements autonomously, eliminating the need for competitor intervention in the verification process.
2Reliability
If human administrators manually identify achievements in video games, then score tracking can occur, but the process is time-consuming and limited in scale
Solution Approach 1:
The patent replaces the slow manual process of human administrators watching and noting achievements with an automated computer vision system that processes livestream videos in real-time. The system uses trained machine learning models to automatically detect events, extract achievement information, and update scores continuously without human intervention, dramatically increasing processing speed while maintaining reliability through consistent automated detection across multiple concurrent competitions.
Solution Approach 2:
The system performs preliminary action by pre-training machine learning models on specific game events and achievement patterns before the actual competition. This preliminary training enables the system to quickly and accurately detect relevant events during live gameplay without requiring real-time human analysis, allowing for rapid score updates while maintaining high accuracy through pre-configured detection capabilities.
3Adaptability or versatility
If multiple administrators are deployed to monitor more competitors, then the number of participants can increase, but operational complexity and costs increase
Solution Approach 1:
The patent implements a universal automated system that can monitor and detect achievements across multiple different video games and competition types simultaneously. The machine learning models are trained to recognize various event types and can adapt to different game environments, allowing a single system to handle diverse competitions without requiring specialized human administrators for each game type. This multi-functionality enables scaling to many participants without proportionally increasing system complexity.
Solution Approach 2:
The patent replaces the complex human organizational structure required to manage multiple administrators with a streamlined automated computer vision system. The system consolidates the functions of multiple human monitors into a single automated platform that can simultaneously track numerous competitors across different games, reducing operational complexity while increasing participant capacity through automated parallel processing of multiple video streams.
Data Source
AI summary
A computing system described herein is configured to obtain a video of a video game being played by a video game player, where an esports competition includes play of the video game by the video game player. The computing system is additionally configured to identify, in the video and through use of machine vision technologies, occurrence of an event in the video game depicted in the video, where the machine vision technologies have been trained to detect occurrences of the event in videos, and further where an outcome of the esports competition is based upon the occurrence of the event in the video game. The computing system is additionally configured to output a value that is indicative of the occurrence of the event in the video game depicted in the video, where a score for the video game player is updated in the esports competition based upon the value that is indicative of the occurrence of the event in the video game depicted in the video.