Symptom Detection System Using Adaptive CNN Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to effectively detect and track individuals exhibiting symptoms of contagious illnesses like coughs and sneezes in public and commercial settings, leading to challenges in resource management and public safety.
Innovation Solution
Integration of audio and video sensors within an IoT system, utilizing convolutional neural networks (CNNs) to detect symptoms and thermopile sensors for temperature data, with adaptive selection of CNN models based on confidence values, and connected lighting systems for visual notifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If audio sensors alone are used to detect symptoms, then the system is simple, but detection accuracy is insufficient especially when audio signals are noisy
Solution Approach 1:
The patent combines audio sensors, video sensors, and thermopile sensors into an integrated symptom detection system. The audio sensor captures cough and sneeze sounds, the video sensor captures visual symptoms, and the thermopile sensor captures thermal data. These multiple sensor types are merged to provide complementary information that improves detection accuracy, especially in noisy environments where audio alone may be insufficient.
Solution Approach 2:
The system employs multi-functional sensors that can detect multiple types of symptoms through different modalities. The same sensor system can detect respiratory symptoms (cough, sneeze), visual symptoms (facial expressions, body movements), and thermal symptoms (temperature changes), making the system universally applicable for detecting various infectious disease symptoms without requiring separate specialized systems.
2Reliability
If multiple CNN models are used with adaptive selection, then detection reliability improves, but processing time and computational complexity increase
Solution Approach 1:
The system dynamically selects which CNN model to apply based on the confidence values from initial audio analysis. When audio signals show high confidence for a specific symptom, the corresponding specialized CNN model is selected for further analysis. This dynamic adaptation allows the system to process high-confidence cases quickly while allocating more computational resources to ambiguous cases, thereby improving overall detection reliability without uniformly increasing processing time for all cases.
Solution Approach 2:
The system applies multiple CNN models selectively rather than always applying all models to every input. By using adaptive selection based on confidence thresholds, the system performs partial analysis with simpler models for clear cases and only engages more complex multiple model analysis when necessary, reducing average processing time while maintaining high detection reliability for challenging cases.
3Measurement precision
If integrated audio and video sensors are deployed, then detection accuracy improves, but device complexity and implementation difficulty increase
Solution Approach 1:
The patent integrates multiple sensor types (audio, video, thermopile) into a unified symptom detection platform that can detect various infectious disease symptoms through different modalities. This multi-functional approach allows a single system to handle diverse detection needs, improving accuracy while consolidating what would otherwise require multiple separate systems, thereby managing implementation complexity.
Solution Approach 2:
The system employs an intermediary processing layer that coordinates data from multiple sensor types and multiple CNN models. This intermediary architecture manages the complexity of integrating audio, video, and thermal data by providing a standardized interface for data fusion and model selection, making the system easier to implement and maintain despite the sophisticated sensor integration.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Accurate and efficient detection and tracking of symptomatic individuals, enabling timely disinfection actions and public notification, reducing the risk of infection spread in crowded environments.
Implementation Method 1
thermopile sensors are integrated in or added to light emitting devices
Implementation Method 2
identifying symptoms using audio signals from microphones
Data Source
AI summary
Systems for detecting and localizing a person exhibiting a symptom of infection in a space are provided. The systems include a user interface configured to receive position information of the space and a plurality of connected sensors in the space, wherein the plurality of connected sensors are configured to capture sensor signals from the person exhibiting the symptom of infection. The systems further include a processor configured to input captured sensor signals from the plurality of connected sensors to at least one convolutional neural network (CNN) model selected based on a confidence value, wherein the processor is further configured to locate the symptomatic person. The systems further include a graphical user interface connected to the processor and configured to display the location of the person exhibiting the symptom of infection within the space.