Apparatus and methods for object detection using machine learning processes
Methods, systems, and apparatuses are provided to automatically detect objects within images. For example, an image capture device may capture an image, and may apply a trained neural network to the image to generate an object value and a class value for each of a plurality of portions of the image. Further, the image capture device may determine, for each of the plurality of image portions, a confidence value based on the object value and the class value corresponding to each image portion. The image capture device may also detect an object within at least one image portion based on the confidence values. Further, the image capture device may output a bounding box corresponding to the at least one image portion. The bounding box defines an area of the image that includes one or more objects.
1. An image capture device comprising:
a non-transitory, machine-readable storage medium storing instructions; and
at least one processor coupled to the non-transitory, machine-readable storage medium, the at least one processor being configured to execute the instructions to:
obtain image data from a camera of the image capture device;
apply a trained machine learning process to the image data to generate an object value and a class value for each of a plurality of image portions of the image data;
determine, for each of the plurality of image portions, a confidence value based on the object value and the class value corresponding to each image portion; and
detect an object within at least one image portion of the plurality of image portions based on the confidence values.
2. The image capture device of claim 1 , wherein the at least one processor is further configured to execute the instructions to:
output a bounding box corresponding to the at least one image portion of the plurality of image portions.
3. The image capture device of claim 2 , wherein the at least one processor is further configured to execute the instructions to:
perform at least one of automatic focus, automatic gain, automatic exposure, and automatic white balance based on the bounding box.
4. The image capture device of claim 1 , wherein detecting the object within the at least one image portion of the plurality of image portions based on the confidence values comprises:
comparing at least one of the confidence values to a threshold; and
determining the object is within the at least one image portion based on the at least one comparison.
5. The image capture device of claim 1 , wherein the trained machine learning process comprises establishing a neural network.
6. The image capture device of claim 5 , wherein the neural network comprises a plurality of convolutional layers, a flattening layer, and a linear layer configured to generate at least one fully connected layer that provides the object value and the class value.
7. The image capture device of claim 1 , wherein the at least one processor is further configured to execute the instructions to:
obtain temporal data characterizing at least one previously generated confidence value; and
apply the trained machine learning process to the temporal data to generate the object value and the class value.
8. The image capture device of claim 7 , wherein the temporal data further characterizes at least one previously generated bounding box.
9. The image capture device of claim 1 , wherein the at least one processor is further configured to execute the instructions to:
apply the trained machine learning process to the image data to generate a keypoint value for each of the plurality of image portions of the image data; and
output the keypoint value for each of the plurality of image portions of the image data.
10. The image capture device of claim 1 , wherein the at least one processor is further configured to execute the instructions to:
apply the trained machine learning process to the image data to generate a visible keypoint value for each of the plurality of image portions of the image data; and
output the visible keypoint value for each of the plurality of image portions of the image data.
11. A method for detecting an object within a captured image, comprising:
obtaining image data from a camera of an image capture device;
applying a trained machine learning process to the image data to generate an object value and a class value for each of a plurality of image portions of the image;
determining, for each of the plurality of image portions, a confidence value based on the object value and the class value corresponding to each image portion; and
detecting the object within at least one image portion of the plurality of image portions based on the confidence values.
12. The method of claim 11 , further comprising:
outputting a bounding box corresponding to the at least one image portion of the plurality of image portions.
13. The method of claim 12 , further comprising:
performing at least one of automatic focus, automatic gain, automatic exposure, and automatic white balance based on the bounding box.
14. The method of claim 11 , wherein the trained machine learning process comprises establishing a neural network.
15. The method of claim 14 , wherein the neural network comprises a plurality of convolutional layers, a flattening layer, and a linear layer configured to generate at least one fully connected layer that provides the object value and the class value.
16. The method of claim 11 , further comprising:
obtaining temporal data characterizing at least one previously generated confidence value; and
applying the trained machine learning process to the temporal data to generate the object value and the class value.
17. The method of claim 11 , further comprising:
applying the trained machine learning process to the image data to generate a keypoint value for each of the plurality of image portions of the image data; and
outputting the keypoint value for each of the plurality of image portions of the image data.
18. The method of claim 11 , further comprising:
applying the trained machine learning process to the image data to generate a visible keypoint value for each of the plurality of image portions of the image data; and
outputting the visible keypoint value for each of the plurality of image portions of the image data.
19. A non-transitory, machine-readable storage medium storing instructions that, when executed by at least one processor, cause the at least one processor to perform operations that include:
obtaining image data from a camera of an image capture device;
applying a trained machine learning process to the image data to generate an object value and a class value for each of a plurality of image portions of the image;
determining, for each of the plurality of image portions, a confidence value based on the object value and the class value corresponding to each image portion; and
detecting an object within at least one image portion of the plurality of image portions based on the confidence values.
20. The non-transitory, machine-readable storage medium of claim 19 , wherein the instructions, when executed by the at least one processor, cause the at least one processor to perform operations including:
outputting a bounding box corresponding to the at least one image portion of the plurality of image portions.