17979256. IMAGE PROCESSING APPARATUS, CONTROL METHOD THEREOF, AND STORAGE MEDIUM simplified abstract (CANON KABUSHIKI KAISHA)
IMAGE PROCESSING APPARATUS, CONTROL METHOD THEREOF, AND STORAGE MEDIUM
Organization Name
Inventor(s)
IMAGE PROCESSING APPARATUS, CONTROL METHOD THEREOF, AND STORAGE MEDIUM - A simplified explanation of the abstract
This abstract first appeared for US patent application 17979256 titled 'IMAGE PROCESSING APPARATUS, CONTROL METHOD THEREOF, AND STORAGE MEDIUM
Simplified Explanation
The abstract describes an image processing apparatus that uses a learning model to classify regions in an input image. If the accuracy of the classification is below a certain threshold, the apparatus subdivides the classification labels and generates new training data to improve the model.
- The apparatus trains a learning model using labeled regions in an input image.
- It uses the trained model to estimate classifications on verification data.
- If the estimation accuracy is below a threshold, it subdivides the classification labels and generates new training data.
- The apparatus then retrains the model using the new training data.
Potential Applications
- Image recognition and classification systems
- Object detection and tracking in videos
- Medical imaging analysis
- Autonomous vehicles and robotics
Problems Solved
- Improves the accuracy of image classification by generating new training data for regions with low estimation accuracy.
- Allows for the refinement of the learning model over time by retraining with new data.
Benefits
- Enhanced accuracy in image classification tasks.
- Adaptability to changing data and improved performance over time.
- Reduction in manual labeling efforts by automatically generating new training data.
Original Abstract Submitted
An image processing apparatus comprises a training unit configured to train a learning model using first training data including a first region, which has been given a first classification label, in an input image; an estimation unit configured to perform estimation using the trained learning model and verification data; a generation unit configured to, in a case where an accuracy of a result of the estimation by the estimation unit is less than or equal to a first threshold, give the first region one of second classification labels, into which the first classification label has been subdivided, and generate second training data including the first region, which has been given the second classification label; and a control unit configured to cause the training unit to perform retraining using the second training data.
