Systems and methods for classifying pet information

1 . A computer-implemented method comprising:

receiving an input image from a client device;

executing a first machine learning model on the input image to determine that the input image comprises a view of fecal matter excreted by an animal and output a bounding box on the input image locating the fecal matter within the input image, wherein the first machine learning model was trained on images containing fecal matter and bounding boxes around the fecal matter and images not containing fecal matter;

executing a second machine learning model on the input image with the bounding box to generate a segmentation mask on the input image, wherein the second machine learning model was trained on isolated images of fecal matter with background removed;

generating, based on the segmentation mask, a segmented image isolating the fecal matter from a background of the fecal matter;

executing a third machine learning model on the segmented image to predict a classification score for the fecal matter, wherein the third machine learning model was trained on images associated with a plurality of classification scores;

executing a fourth machine learning model on the segmented image to predict characteristics of the fecal matter, wherein the fourth machine learning model was trained on images containing fecal matter associated with a plurality of characteristics;

determining a microbiome age status of the animal based on correlating one or more bacterial taxa detected in the fecal matter to a control data set;

generating a health assessment of the animal based on the classification score and characteristics of the fecal matter and the microbiome age status;

generating a recommendation for the animal based on the health assessment of the animal; and

displaying information related to the recommendation for the animal to a user.

2 . The computer-implemented method of claim 1 , wherein the classification score of the fecal matter indicates a health metric related to the animal.

3 . The computer-implemented method of claim 1 , further comprising generating and displaying an indication of the confidence score, wherein the confidence score is associated with the classification score of the fecal matter or the health assessment of the animal.

4 . The computer-implemented method of claim 1 , wherein the characteristics of the fecal matter comprises a presence of blood in the fecal matter.

5 . The computer-implemented method of claim 4 , wherein the health assessment is at least partly based on the presence of blood.

6 . The computer-implemented method of claim 1 , wherein the health assessment is at least partly based on at least one of a consistency, a texture, a color, a size, a shape, a volume, a presence of solids, a water content, or a presence of inorganic material of the fecal matter associated with the characteristics of the fecal matter.

7 . The computer-implemented method of claim 1 , wherein the recommendation comprises a recommendation for one or more pet products.

8 . The computer-implemented method of claim 1 , wherein the health assessment is generated partly based on one or more additional inputs, the one or more additional inputs comprising at least one of an activity level of the animal, the animal's diet, or a microbiome analyses of a sample of the animal's fecal matter.

9 . The computer-implemented method of claim 8 , further comprising receiving a digital input indicating the activity level of the pet from a tracker worn by the pet over a period of time, the tracker comprising at least one of a sensor or an accelerometer.

10 . The computer-implemented method of claim 8 , wherein the recommendation comprises a recommendation for at least one of a pet service, a change in diet, or an exercise routine.

11 . The computer-implemented method of claim 8 , wherein the recommendation comprises a recommendation to take the pet to a veterinarian.

12 . The computer-implemented method of claim 1 , wherein the health assessment is generated partly based on one or more additional inputs, the one or more additional inputs comprising at least one of a species or a breed of the animal.

13 . The computer-implemented method of claim 12 , wherein one or more of the first, second, third, or fourth machine learning model are selected from a set of available machine learning models based on the one or more additional inputs.

14 . The computer-implemented method of claim 1 , wherein the health assessment is generated partly based on one or more additional inputs, the one or more additional inputs comprising at least one of an age, a gender, a breed, a weight, or a height of the animal.

15 . The computer-implemented method of claim 1 , wherein the health assessment is generated partly based on one or more additional inputs, the one or more additional inputs comprising a calculated biological age of the animal.

16 . One or more computer-readable non-transitory storage media embodying software that is operable when executed to:

receive an input image from a client device;

execute a first machine learning model on the input image to determine that the input image comprises a view of fecal matter excreted by an animal and output a bounding box on the input image locating the fecal matter within the input image, wherein the first machine learning model was trained on images containing fecal matter and bounding boxes around the fecal matter and images not containing fecal matter;

execute a second machine learning model on the input image with the bounding box to generate a segmentation mask on the input image, wherein the second machine learning model was trained on isolated images of fecal matter with background removed;

generate, based on the segmentation mask, a segmented image isolating the fecal matter from a background of the fecal matter;

execute a third machine learning model on the segmented image to predict a classification score for the fecal matter, wherein the third machine learning model was trained on images associated with a plurality of classification scores;

execute a fourth machine learning model on the segmented image to predict characteristics of the fecal matter, wherein the fourth machine learning model was trained on images containing fecal matter associated with a plurality of characteristics;

determine a microbiome age status of the animal based on correlating one or more bacterial taxa detected in the fecal matter to a control data set;

generate a health assessment of the animal based on the classification score and characteristics of the fecal matter and the microbiome age status;

generate a recommendation for the animal based on the health assessment of the animal; and

display information related to the recommendation for the animal to a user.

17 . The storage media of claim 16 , wherein the classification score of the fecal matter indicates a health metric related to the animal.

18 . A system comprising:

one or more processors; and

one or more computer-readable non-transitory storage media coupled to one or more of the processors and comprising instructions operable when executed by one or more of the processors to cause the system to:

execute a first machine learning model on the input image to determine that the input image comprises a view of fecal matter excreted by an animal and output a bounding box on the input image locating the fecal matter within the input image, wherein the first machine learning model was trained on images containing fecal matter and bounding boxes around the fecal matter and images not containing fecal matter;

execute a second machine learning model on the input image with the bounding box to generate a segmentation mask on the input image, wherein the second machine learning model was trained on isolated images of fecal matter with background removed;

generate, based on the segmentation mask, a segmented image isolating the fecal matter from a background of the fecal matter;

execute a third machine learning model on the segmented image to predict a classification score for the fecal matter, wherein the third machine learning model was trained on images associated with a plurality of classification scores;

execute a fourth machine learning model on the segmented image to predict characteristics of the fecal matter, wherein the fourth machine learning model was trained on images containing fecal matter associated with a plurality of characteristics;

determine a microbiome age status of the animal based on correlating one or more bacterial taxa detected in the fecal matter to a control data set;

generate a health assessment of the animal based on the classification score and characteristics of the fecal matter and the microbiome age status;

generate a recommendation for the animal based on the health assessment of the animal; and

display information related to the recommendation for the animal to a user.