Machine-Learned Lighting Feature Generation from Display Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lighting control systems for performances lack the ability to create complex light shows that adequately account for the specific music being played and the available lighting technology, requiring significant manual effort and failing to provide deep integration.
Innovation Solution
A method utilizing a machine-learning model trained on performance and lighting data to generate new lighting features and information, analyzing relationships between music and lighting to automate light show creation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If complex lighting control sequences are created to achieve close coordination between music and light show, then the quality of light show coordination is improved, but the workload and complexity for operating personnel increases
Solution Approach 1:
The patent replaces manual mechanical control processes with an automated system that uses machine learning models, audio analysis, and performance data processing to generate lighting control sequences automatically, eliminating the need for operators to manually create complex control sequences
Solution Approach 2:
The system enables self-service by automatically analyzing performance data and generating appropriate lighting control sequences without requiring expert operators to manually program the sequences, allowing the system to serve itself in creating coordinated light shows
2Extent of automation
If existing lighting control methods are used that only rudimentarily analyze music, then the device complexity is reduced, but the light show coordination quality and automation extent deteriorates
Solution Approach 1:
The system performs preliminary actions by pre-processing performance data, training machine learning models in advance, and preparing analysis frameworks before actual light show generation, enabling automated high-quality coordination without increasing operational complexity during execution
Solution Approach 2:
The patent introduces intermediary components including machine learning models, audio analysis modules, and performance data processing systems that mediate between the music/performance input and lighting control output, enabling sophisticated automation while managing system complexity through modular architecture
3Measurement precision
If more training data sets are used to improve machine learning model accuracy, then the light show generation quality is improved, but the data processing time and computational resources increase
Solution Approach 1:
The system applies partial action by using a representative subset of training data that provides sufficient accuracy for light show generation without processing every possible data point, achieving adequate precision while limiting training time and computational resources
Solution Approach 2:
The patent employs parameter changes by adjusting machine learning model parameters, data sampling rates, and training iteration counts to optimize the balance between accuracy and training time, finding optimal settings that achieve sufficient light show generation quality without excessive computational cost
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The invention relates to a method for generating at least one new piece of illumination information (8) and/or at least one new illumination feature, wherein at least one training data set comprising mutually associated presentation information (1) and illumination information (2) is obtained. Furthermore, a machine-learning model (5) is created using mutually associated illumination features (4) and presentation features (3), wherein the illumination features (4) and presentation features (3) are obtained by analysis before and/or during the creation of the machine-learning model (5).Subsequently, new presentation information (6) is obtained, and at least one new illumination feature (8) and/or at least one new illumination information item is generated by inference with the machine-learning model (5) upon input of the new presentation information (6) and/or the new presentation features (7) obtained by analyzing the new presentation information (6). The invention further relates to a computer system (20) configured to carry out one of the methods, and to a computer program product comprising software code sections configured such that the method can be executed by at least one processor (24).