Object detection using multiple neural network configurations
AssigneeWestern Digital Technologies, Inc.
InventorsHaoyu WU, Shaomin XIONG, Toshiki HIRANO
Life of this patent
Estimating the term…
Abstract
This disclosure relates to an apparatus for object detection. The apparatus comprises a video camera, an object detector, and a controller. The video camera may be configured to generate a video stream of frames. The object detector may be configured to accept the video stream as input data and to perform object detection. The controller may be coupled to the video camera and the object detector. The controller may be configured to manage object detection in order to satisfy a performance metric and/or operate within an operational constraint.
Claims20
An apparatus, comprising:
- a repository storing a plurality of object detectors;
- an object detector selected from the plurality of object detectors and configured to accept a stream of video frames as input data and to perform object detection based on the input data to generate an object detection result;
- a controller coupled to the object detector and configured to manage the object detection in order to satisfy one or more of a performance metric and an operational constraint;
- a baseline object detector coupled to the controller and configured to generate a baseline object detection result;
- a comparator configured to:
- measure the object detection result against the baseline object detection result;
- and
- generate a performance score for the object detector;
- and
- a detector selector coupled to the comparator and configured to select a new object detector from the plurality of object detectors in the repository based on the performance score.
A method, comprising:
- selecting an object detector from a plurality of object detectors stored in a repository;
- processing a stream of video frames using the object detector to perform object detection and to generate an object detection result;
- managing the object detection with a controller coupled to the object detector in order to satisfy one or more of a performance metric and an operational constraint;
- generating a baseline object detection result using a baseline object detector coupled to the controller;
- comparing the object detection result with the baseline object detection result to generate a performance score for the object detector using a comparator coupled to the controller;
- and
- selecting a new object detector from the plurality of object detectors in the repository based on the performance score using a detector selector coupled to the comparator.
Dependencies are inferred from the claim text.