Dynamic object classification
First Claim
1. A method of improving performance of video analytics for a camera system in response to a detection preference of a system user, comprising:
- while the camera system is already in in-field use, receiving image data representing multiple images of a scene of a field of view of the camera system, the multiple images including representations of multiple objects, a first set of the multiple objects having members of an object class, and a second set of the multiple objects not having members of the object class;
while the camera system is already in in-field use, using video analytics implemented with a general classifier that performs general classifier steps in analyzing the received image data to produce a general classification determination classifying the multiple objects as either members or non-members of the object class;
while the camera system is already in in-field use, generating mistake metadata in response to acknowledgement by the system user that the general classification determination resulted in a mistaken classification determination based on the detection preference of the system user; and
while the camera system is already in in-field use, improving video analytics performance based on the mistake metadata by performing a specialization step in addition to the general classification steps performed, the specialization step producing for presentation to the system user a specialized classification determination of the multiple objects classified by the general classifier so as to reduce a number of future mistaken classification determinations.
7 Assignments
0 Petitions
Accused Products
Abstract
A camera system comprises an image capturing device and an object classification module connected to the image capturing device. The image capturing device has a field of view and produces image data representing an image of the field of view. The object classification module is operable to determine whether an object in an image is a member of an object class. The object classification module includes N decision steps configured in a cascade configuration, wherein at least one of the N decision steps is operable to (a) accept an object as a member of the object class, (b) reject an object as a member of the object class, and (c) call on a next step to determine whether an object is a member of the object class.
-
Citations
30 Claims
-
1. A method of improving performance of video analytics for a camera system in response to a detection preference of a system user, comprising:
-
while the camera system is already in in-field use, receiving image data representing multiple images of a scene of a field of view of the camera system, the multiple images including representations of multiple objects, a first set of the multiple objects having members of an object class, and a second set of the multiple objects not having members of the object class; while the camera system is already in in-field use, using video analytics implemented with a general classifier that performs general classifier steps in analyzing the received image data to produce a general classification determination classifying the multiple objects as either members or non-members of the object class; while the camera system is already in in-field use, generating mistake metadata in response to acknowledgement by the system user that the general classification determination resulted in a mistaken classification determination based on the detection preference of the system user; and while the camera system is already in in-field use, improving video analytics performance based on the mistake metadata by performing a specialization step in addition to the general classification steps performed, the specialization step producing for presentation to the system user a specialized classification determination of the multiple objects classified by the general classifier so as to reduce a number of future mistaken classification determinations. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
18. A field-deployed camera system, comprising:
-
an object classification module operable, while the camera system is already in in-field use, to receive image data including representations of objects captured in images of a field of view of an image capturing device, the object classification module including an object classifier that is operable to determine whether the objects in the images are members of an object class, wherein the object classifier generates mistaken classifications; a user station having a display for presenting the images of the objects to a user, the user station being operable to present on the display representations of the mistaken classifications generated by the object classifier, wherein the user station is operable to produce user feedback information in response to user acknowledgement of the mistaken classifications, the user feedback generating mistake metadata; and a classifier evolution module receiving the mistake metadata and being operable, while the camera system is already in in-field use, to modify the object classifier using the mistake metadata so as to reduce the number of future mistaken classifications, the classifier evolution module thereby generating a specialized classifier. - View Dependent Claims (19, 20, 21, 22, 23)
-
-
24. A method of improving performance of video analytics for a camera system in response to a detection preference of a system user, comprising:
-
while the camera system is already in in-field use, receiving image data representing multiple images of a scene of a field of view of the camera system, the multiple images including representations of multiple objects, a first set of the multiple objects having members of an object class, and a second set of the multiple objects not having members of the object class; while the camera system is already in in-field use, using video analytics implemented with a general classifier that performs general classifier steps in analyzing the received image data to produce a general classification determination classifying the multiple objects as either members or non-members of the object class; while the camera system is already in in-field use, generating a user feedback indication in response to acknowledgement by the system user that the general classification determination resulted in a mistaken classification determination based on detection preference of the system user; and while the camera system is already in in-field use, using the user feedback indication and video analytics implemented with the general classifier to perform the general classification steps in analyzing the received image data to construct a feedback-based classifier that performs classification based on the system user feedback; while the camera system is already in in-field use, testing operational performance of the feedback-based classifier in analyzing the received image data to produce a feedback-based classification determination classifying the multiple objects as either members or non-members of the object class; and while the camera system is already in in-field use, evaluating whether the feedback-based classification determination in comparison with the general classification determination is more consistent with the detection preference of the system user and thereby validates an improved performance of the video analytics. - View Dependent Claims (25, 26)
-
-
27. A method of improving performance of video analytics for a camera system in response to a detection preference of a system user, comprising:
-
while the camera system is already in in-field use, receiving image data representing multiple images of a scene of a field of view of the camera system, the multiple images including representations of multiple objects, a first set of the multiple objects having members of an object class, and a second set of the multiple objects not having members of the object class; while the camera system is already in in-field use, using video analytics implemented with a general classifier that performs general classifier steps in analyzing the received image data to produce a general classification determination classifying the multiple objects as either members or non-members of the object class; while the camera system is already in in-field use, generating mistake metadata in response to acknowledgement by the system user that the general classification determination resulted in a mistaken classification determination based on the detection preference of the system user; and while the camera system is already in in-field use, improving video analytics performance based on the mistake metadata by performing a specialization step in addition to the general classification steps performed, the specialization step including labeling the received image data that does not include the mistaken classification as correctly classified image data, labeling the received image data that does the mistaken classification as incorrectly classified image data, generating a new general classifier that is only trained with the correctly classified image data, and replacing the general classifier with the new general classifier. - View Dependent Claims (28, 29, 30)
-
Specification