17850545. METHOD, DEVICE, AND COMPUTER PROGRAM PRODUCT FOR TRAINING IMAGE CLASSIFICATION MODEL simplified abstract (Dell Products L.P.)
METHOD, DEVICE, AND COMPUTER PROGRAM PRODUCT FOR TRAINING IMAGE CLASSIFICATION MODEL
Organization Name
Inventor(s)
METHOD, DEVICE, AND COMPUTER PROGRAM PRODUCT FOR TRAINING IMAGE CLASSIFICATION MODEL - A simplified explanation of the abstract
This abstract first appeared for US patent application 17850545 titled 'METHOD, DEVICE, AND COMPUTER PROGRAM PRODUCT FOR TRAINING IMAGE CLASSIFICATION MODEL
Simplified Explanation
The abstract describes a method, device, and computer program for training an image classification model. Here are the key points:
- The method involves training an image classification model to classify pixel points of a sample image into different object classes.
- In each training iteration, the model determines a classification result for the pixel points.
- The first classification result is obtained in the first training iteration, and the second classification result is obtained in the second training iteration.
- Based on these classification results, the method determines recall rates for each object class in both iterations.
- The recall rates represent the model's ability to correctly identify instances of each object class.
- The image classification model is then adjusted based on these recall rates to improve its performance.
- The result is a trained image classification model that can accurately classify objects in images.
Potential applications of this technology:
- Image recognition and classification systems
- Object detection in images and videos
- Autonomous vehicles for identifying and tracking objects on the road
- Surveillance systems for detecting specific objects or individuals
Problems solved by this technology:
- Improves the accuracy and performance of image classification models
- Enables more precise identification and classification of objects in images
- Reduces false positives and false negatives in object detection systems
Benefits of this technology:
- Enhanced accuracy in identifying and classifying objects in images
- Improved performance of image classification models
- Increased reliability and efficiency in object detection systems
- Enables the development of more advanced and reliable computer vision applications.
Original Abstract Submitted
Embodiments disclosed herein relate to a method, a device, and a computer program product for training an image classification model. The method includes: determining a first classification result obtained by the image classification model on pixel points of a sample image in a first training iteration, wherein the first classification result indicates that each of the pixel points belongs to one of a plurality of object classes; determining a second classification result obtained by the image classification model on the pixel points of the sample image in a second training iteration; determining, based on the first classification result and the second classification result, a first set of recall rates and a second set of recall rates for the plurality of object classes; and adjusting, based on the first set of recall rates and the second set of recall rates, the image classification model to obtain a trained image classification model.
