Systems and methods for image object recognition based on location information and object categories
Systems, methods, and non-transitory computer-readable media can identify a set of regions corresponding to a geographical area. A collection of training images can be acquired. Each training image in the collection can be associated with one or more respective recognized objects and with a respective region in the set of regions. Histogram metrics for a plurality of object categories within each region in the set of regions can be determined based at least in part on the collection of training images. A neural network can be developed based at least in part on the histogram metrics for the plurality of object categories within each region in the set of regions and on the collection of training images.
1. A computer-implemented method comprising:
developing, by a computing system, a neural network based at least in part on training data including training images and metrics for a plurality of object categories associated with geographical regions in a set of geographical regions based on the training images;
inputting, by the computing system, images into the neural network; and
generating, by the computing system, object recognition outputs associated with the images based on the neural network.
2. The computer-implemented method of claim 1 , wherein the training data further comprises the set of geographical regions associated with the training images and recognized objects associated with the training images.
3. The computer-implemented method of claim 2 , wherein the recognized objects are determined from at least one of an image classification process applied to each training image or one or more hashtags associated with each training image.
4. The computer-implemented method of claim 2 , wherein a geographical region in the set of geographical regions associated with each training image is determined based at least in part on acquisition of location information for each training image and identification of the geographical region associated with the location information.
5. The computer-implemented method of claim 1 , wherein the developing the neural network comprises:
generating a table representing metrics for a plurality of object categories associated with a geographical region in the set of geographical regions;
providing a location layer within the neural network, the location layer being associated with the table; and
training the neural network based at least in part on the training data.
6. The computer-implemented method of claim 5 ,
wherein the inputting images into the neural network comprises inputting a particular image and location information associated with the particular image into the neural network; and
wherein the generating object recognition outputs comprises causing the neural network to produce, based at least in part on the location information associated with the particular image, an object recognition output for the particular image.
7. The computer-implemented method of claim 5 , further comprising:
determining, based at least in part on the training the neural network, a preferred radius associated with an object category in a plurality of object categories.
8. The computer-implemented method of claim 1 , wherein a geographical region in the set of geographical regions is associated with a portion of the neural network via a connection in a group of connections, and wherein the connection in the group of connections has a weight.
9. The computer-implemented method of claim 8 , wherein the developing the neural network comprises training the neural network so that the weight of the connection is modified to produce an object recognition output for a particular training image that is closer to one or more recognized objects associated with the particular training image.
10. The computer-implemented method of claim 1 ,
wherein the metrics are a histogram count associated with the plurality of object categories.
11. A system comprising:
at least one processor; and
a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
developing a neural network based at least in part on training data including training images and metrics for a plurality of object categories associated with geographical regions in a set of geographical regions based on the training images;
inputting images into the neural network; and
generating object recognition outputs associated with the images based on the neural network.
12. The system of claim 11 , wherein the training data further comprises the set of geographical regions associated with the training images and recognized objects associated with the training images.
13. The system of claim 12 , wherein the recognized objects are determined from at least one of an image classification process applied to each training image or one or more respective hashtags associated with each training image.
14. The system of claim 12 , wherein a geographical region in the set of geographical regions associated with each training image is determined based at least in part on acquisition of location information for each training image and identification of the geographical region associated with the location information.
15. The system of claim 11 , wherein the developing the neural network comprises:
generating a table representing metrics for a plurality of object categories associated with a geographical region in the set of geographical regions;
providing a location layer within the neural network, the location layer being associated with the table; and
training the neural network based at least in part on the training data.
16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to perform:
developing a neural network based at least in part on training data including training images and metrics for a plurality of object categories associated with geographical regions in a set of geographical regions based on the training images;
inputting images into the neural network; and
generating object recognition outputs associated with the images based on the neural network.
17. The non-transitory computer-readable storage medium of claim 16 , wherein the training data further comprises the set of geographical regions associated with the training images and recognized objects associated with the training images.
18. The non-transitory computer-readable storage medium of claim 17 , wherein the recognized objects are determined from at least one of an image classification process applied to each training image or one or more respective hashtags associated with each training image.
19. The non-transitory computer-readable storage medium of claim 17 , wherein a geographical region in the set of geographical regions associated with each training image is determined based at least in part on acquisition of location information for each training image and identification of the geographical region associated with the location information.
20. The non-transitory computer-readable storage medium of claim 16 , wherein the developing the neural network comprises:
generating a table representing metrics for a plurality of object categories associated with a geographical region in the set of geographical regions;
providing a location layer within the neural network, the location layer being associated with the table; and
training the neural network based at least in part on the training data.