A train-time loss in a system and method for calibrating object detection
Munir, Muhammad Akhtar Khan, Muhammad Haris Khan, Salman Khan, Fahad Shahbaz
Munir, Muhammad Akhtar
Khan, Muhammad Haris
Khan, Salman
Khan, Fahad Shahbaz
Supervisor
Department
Computer Vision
Embargo End Date
Type
Patent
Date
2025
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
A system and method of training a deep neural network for object detection in an object detection system. The object detection system including a camera and a controller including the DNN. The method including capturing an image by the camera, receiving the image, predicting, using the DNN, a bounding box and corresponding class label, evaluating the prediction with a total loss function including an object detection loss function, a box regression loss function, and a calibration loss function that takes into account precision and confidence. The method outputs a calibrated image with the object bounding box, the corresponding label, and a respective confidence score, in which the confidence score is a probability associated with the predicted class label.
Citation
A Train-Time Loss in A System and Method for Calibrating Object Detection, by M. A. Munir, M. H. Khan, S. Khan, F. S. Khan (2025, Feb. 20). Patent US20250061697A1 [Online]. Available: https://patents.google.com/patent/US20250061697A1/en
Source
US Patent App. 18/526,254
Conference
Keywords
Object detection calibration, Deep neural networks, Precision-confidence alignment, Train-time loss function, Safety-critical applications
Subjects
Source
Publisher
Google Patent