Smiling Face Detection Using Classifier and Clutter Filter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing camera systems face challenges in accurately detecting and tracking smiling faces, particularly in distinguishing between genuine smiling faces and other objects, leading to potential false detections and suboptimal image capture.
Innovation Solution
A method and apparatus for smiling face detection using an object classifier trained to differentiate between smiling faces and non-smiling faces, employing a combination of image processing units for real-time detection and tracking, with a clutter filter to enhance confidence levels and reduce false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a standard object classifier is used for face detection, then the detection speed is fast, but the accuracy in distinguishing smiling faces from non-smiling faces is low
Solution Approach 1:
The system performs preliminary actions by pre-processing images to extract facial features and expressions before classification. The object classifier is trained in advance with labeled datasets of smiling and non-smiling faces, enabling accurate distinction without requiring complex real-time analysis during image capture.
Solution Approach 2:
The detection process is segmented into multiple stages: image capture, face detection, facial feature extraction, expression analysis, and classification. This segmentation allows each component to be optimized independently, with the classifier focusing solely on distinguishing smiling expressions from non-smiling ones using pre-extracted features.
2Speed
If real-time smiling face detection is implemented, then the response time is reduced, but the false detection rate increases
Solution Approach 1:
The system incorporates feedback mechanisms where detected smiling faces are verified against tracked face lists from previous frames. The clutter filter uses feedback from multiple detection rounds and temporal information to confirm whether a detected face is a genuine smile, reducing false positives while maintaining real-time response.
Solution Approach 2:
The detection system dynamically adjusts its sensitivity and filtering parameters based on the scene context and detection history. The object classifier and clutter filter work together to adaptively distinguish between genuine smiles and false detections, with the system learning from previous frames to improve accuracy over time while maintaining speed.
3Measurement precision
If a clutter filter is added to reduce false positives, then the detection accuracy improves, but the processing time increases
Solution Approach 1:
The clutter filter performs preliminary filtering actions by comparing detected faces against a pre-maintained list of tracked faces from previous frames. This pre-computed tracking information is used to quickly identify and reject false detections without requiring extensive real-time analysis, thus improving accuracy with minimal time penalty.
Solution Approach 2:
The system applies partial filtering actions by only performing thorough verification on detections that pass initial quick checks. The clutter filter uses a multi-stage approach with progressively stricter criteria, applying full analysis only when necessary, which balances accuracy improvement with processing time constraints.
Data Source
AI summary
A method, apparatus, and manufacture for smiling face detection is provided. For each frame, a list of new smiling faces for the frame is generated by performing smiling face detection employing an object classifier that trained is to distinguish between smiling faces and all objects in the frame that are not smiling faces. For the first frame, the list of new smiling faces is employed as an input smiling face list for the next frame. For each frame after the first frame, a list of tracked smiles for the frame is generated by tracking smiling faces in the frame from the input smiling list for the frame. Further, a list of new smiling faces is generated for the next frame by combining the list of new smiling faces for the frame with the list of tracked smiles for the frame.