Predictive classification of insects
AssigneeVerily Life Sciences LLC
InventorsMark Desnoyer, Victor Criswell, Josh Livni, Yaniv Ovadia, Peter Massaro, Nigel Snoad, Dilip Krishnan, Yi Han, Tiantian Zha
Life of this patent
Estimating the term…
Abstract
Insects can be classified into a category (e.g., sex category, species category, size category, etc.) using a variety of different classification approaches including, for example, an industrial vision classifier and/or a machine learning classifier. At least some classification approaches may be used in real-time to make real-time decisions and others can be used to validate earlier-made real-time decisions.
Claims35
A system, comprising:
- an imaging device configured to capture images of insects;
- and
- a computing device in communication with the imaging device, and configured to at least:
- instruct the imaging device to capture an image that depicts at least a portion of the insect;
- receive, from the imaging device, the image;
- determine, using an industrial vision classifier:
- (i) a first classification of the image into at least one category based at least in part on features extracted from the image;
- and
- (ii) a first confidence measure corresponding to the first classification;
- determine, using a machine learning classifier:
- (i) a second classification of the image into the at least one category based at least in part on the image;
- and
- (ii) a second confidence measure corresponding to the second classification;
- and
- determine a third classification of the image based at least in part on the first confidence measure and the second confidence measure.
A computer-implemented method, comprising:
- receiving a sequence of images that each depict at least a portion of an insect;
- determining, using an industrial vision classifier, a first classification of the sequence of images insect into at least one category;
- determining a first confidence measure corresponding to the first classification;
- in the event the first confidence measure falls below a threshold, determining, using a machine learning classifier, a second classification of the sequence of images into the at least one category;
- determining a second confidence measure corresponding to the second classification;
- and
- generating classification information relating to the insect based on the second confidence measure.
A system, comprising:
- an imaging device configured to capture images of insects;
- and
- a computing device in communication with the imaging device, and configured to at least:
- instruct the imaging device to capture a sequence of images depicting at least a portion of an insect;
- use a first predictive model to determine a first output corresponding to a first classification of a first image of the sequence of images, the first output comprising a confidence measure of the first classification;
- and
- generate classification information based at least in part on the first output.
One or more non-transitory computer-readable media comprising computer-executable instructions that, when executed by one or more computing systems, cause the one or more computing systems to:
- receive a set of images depicting at least a portion of an insect;
- determine a first set of classifications for the set of images by at least inputting the set of images into a first predictive model, the first predictive model comprising a core deep neural network model;
- determine a first set of confidence measures for the first set of classifications;
- determine a second set of classifications for the set of images by at least inputting the first set of confidence measures into a second predictive model, the second predictive model comprising a recurrent neural network model;
- determine a second set of confidence measures for the second set of classifications;
- and
- generate classification information based at least in part on the second set of confidence measures.
A computer-implemented method, comprising:
- receiving a set of images depicting at least a portion of an insect;
- determining a first classification for the set of images by least inputting the set of images into a first predictive model, the first predictive model comprising a core deep neural network model;
- determining a set of features corresponding to the first classification;
- determining a second classification for the set of images by at least inputting the set of features into a second predictive model, the second predictive model comprising a recurrent neural network model;
- determining a second confidence measure for the second classification;
- and
- generating classification information based at least in part on the second confidence measure.
Dependencies are inferred from the claim text.