17457264. OBJECT DETECTION CONSIDERING TENDENCY OF OBJECT LOCATION simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

OBJECT DETECTION CONSIDERING TENDENCY OF OBJECT LOCATION

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

HIROKI Kawasaki of Kanagawa (JP)

Shingo Nagai of Tokyo (JP)

OBJECT DETECTION CONSIDERING TENDENCY OF OBJECT LOCATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 17457264 titled 'OBJECT DETECTION CONSIDERING TENDENCY OF OBJECT LOCATION

Simplified Explanation

The patent application describes a method, computer system, and computer program for object detection. Here are the key points:

  • The method involves receiving an annotated image dataset with rectangles surrounding objects and labels specifying their class.
  • It calculates areas of high and low probability of rectangle distribution for each class of objects within the dataset.
  • A correction factor is applied to confidence values of object prediction results based on the class label and rectangle location.
  • The accuracy of the trained object detection (OD) model is calculated, and the correction factor is increased and accuracy re-calculated iteratively.
  • The optimal correction factor is selected, which yields the highest accuracy for the trained OD model.

Potential applications of this technology:

  • Object detection in various fields such as autonomous vehicles, surveillance systems, and robotics.
  • Enhancing the accuracy of object detection models used in image recognition and computer vision tasks.

Problems solved by this technology:

  • Improves the accuracy of object detection models by considering the distribution of rectangles and class labels.
  • Provides a systematic approach to adjusting confidence values based on object location and class, leading to more reliable predictions.

Benefits of this technology:

  • Increases the accuracy of object detection models, resulting in more precise identification and localization of objects.
  • Reduces false positives and false negatives in object detection, improving the overall performance of computer vision systems.
  • Enables better decision-making in applications like autonomous vehicles and surveillance systems, enhancing safety and security.


Original Abstract Submitted

According to one embodiment, a method, computer system, and computer program product for object detection. The embodiment may include receiving an annotated image dataset comprising rectangles which surround objects to be detected and labels which specify a class to which an object belongs. The embodiment may include calculating areas of high and low probability of rectangle distribution for each class of objects within images of the dataset. The embodiment may include applying a correction factor to confidence values of object prediction results, obtained during validation of a trained object detection (OD) model, depending on a class label and a rectangle location of an object prediction result and calculating an accuracy of the trained OD model. The embodiment may include increasing the correction factor and re-calculating the accuracy of the trained OD model with every increase. The embodiment may include selecting an optimal correction factor which yields a highest accuracy.

Retrieved from "http://wikipatents.org/index.php?title=17457264._OBJECT_DETECTION_CONSIDERING_TENDENCY_OF_OBJECT_LOCATION_simplified_abstract_(INTERNATIONAL_BUSINESS_MACHINES_CORPORATION)&oldid=25772"