Sensor data segmentation
A system may include one or more processors configured to receive a plurality of images representing an environment. The images may include image data generated by an image capture device. The processors may also be configured to transmit the image data to an image segmentation network configured to segment the images. The processors may also be configured to receive sensor data associated with the environment including sensor data generated by a sensor of a type different than an image capture device. The processors may be configured to associate the sensor data with segmented images to create a training dataset. The processors may be configured to transmit the training dataset to a machine learning network configured to run a sensor data segmentation model, and train the sensor data segmentation model using the training dataset, such that the sensor data segmentation model is configured to segment sensor data.
1. A system comprising:
one or more processors; and
one or more computer readable storage media communicatively coupled to the one or more processors and storing instructions executable by the one or more processors to:
receive image data;
store, based at least in part on the image data, a segmented image;
receive sensor data from at least one of a light detection and ranging (LIDAR) sensor, a radio detection and ranging (RADAR) sensor, or a sound navigation and ranging (SONAR) sensor;
associate the sensor data with the segmented image as part of a training dataset; and
train a sensor data segmentation model using the training dataset such that, upon receiving input comprising at least one of additional LIDAR sensor data, additional RADAR sensor data, or additional SONAR sensor data, the sensor data segmentation model is configured to output at least one of segmented additional LIDAR sensor data, segmented additional RADAR sensor data, or segmented additional SONAR sensor data.
2. The system of claim 1 , wherein associating the sensor data with the segmented image comprises:
determining a location in the segmented image that corresponds to a portion of the sensor data;
determining, based at least in part on the segmented image and the location, a classification; and
associating the classification with the portion of the sensor data.
3. The system of claim 1 , wherein:
a first portion of the image data is captured at a first image time;
a second portion of image data is captured at a second image time;
a first portion of the sensor data is associated with a first sensor time;
a second portion of the sensor data is associated with a second sensor time; and
the instructions are further executable by the one or more processors to:
determine a first time difference between the first image time and the first sensor time;
determine a second time difference between the second image time and the first sensor time;
determine that the first time difference is less than the second time difference; and
associate the first portion of the first sensor data with the first portion of the image data based at least in part on the first time difference being less than the second time difference.
4. The system of claim 1 , wherein the instructions are further executable by the one or more processors to receive second sensor data and segment the second sensor data based at least in part on the sensor data segmentation model.
5. The system of claim 1 , wherein:
associating the sensor data with the segmented image comprises applying an operation to the sensor data in accordance with a first stride along a first direction and a second stride, different from the first stride, along a second direction, and
training the sensor data segmentation model comprises computing a loss, the loss based at least in part on an output of the sensor data segmentation model and the training dataset, the loss comprising one or more of a cross-entropy softmax loss, a focal loss, or a logistic regression loss.
6. The system of claim 1 , wherein the instructions are further executable by the one or more processors to:
identify discontinuities in the sensor data; and
disregard sensor data associated with the discontinuities.
7. The system of claim 1 , wherein the sensor data comprises LIDAR sensor data from a LIDAR sensor, and wherein the instructions further cause the one or more processors to:
segment, as segmented LIDAR sensor data, the LIDAR sensor data received from the LIDAR sensor using the sensor data segmentation model; and
store the segmented LIDAR sensor data.
8. The system of claim 7 , wherein the instructions are further executable by the one or more processors to transmit the sensor data segmentation model to an autonomous vehicle, the autonomous vehicle configured to generate a trajectory based at least in part on an additional output of the sensor data segmentation model.
9. A method comprising:
receiving image data;
receiving first sensor data generated by at least one of a light detection and ranging (LIDAR) sensor, a radio detection and ranging (RADAR) sensor, or a sound navigation and ranging (SONAR) sensor;
segmenting the image data to store segmented images;
associating the first sensor data with the segmented images to create a training dataset; and
training a sensor data segmentation model using the training dataset such that, upon receiving input comprising at least one of additional LIDAR sensor data, additional RADAR sensor data, or additional SONAR sensor data, the sensor data segmentation model is configured to output at least one of segmented additional LIDAR sensor data, segmented additional RADAR sensor data, or segmented additional SONAR sensor data.
10. The method of claim 9 , wherein receiving the first sensor data comprises receiving sensor data generated by a light detection and ranging (LIDAR) sensor.
11. The method of claim 9 , wherein associating the first sensor data with the segmented images comprises:
determining a location in a segmented image of the segmented images that corresponds to a portion of the first sensor data;
determining, based at least in part on the segmented image and the location, a classification; and
associating the classification with the portion of the first sensor data.
12. The method of claim 9 , wherein:
a first portion of the image data is captured at a first image time;
a second portion of image data is captured at a second image time;
a first portion of the first sensor data is associated with a first sensor time;
a second portion of the first sensor data is associated with a second sensor time; and
the method comprises:
determining a first time difference between the first image time and the first sensor time;
determining a second time difference between the second image time and the first sensor time;
determining that the first time difference is less than the second time difference; and
associating the first portion of the first sensor data with the first portion of the image data based at least in part on the first time difference being less than the second time difference.
13. The method of claim 9 , further comprising:
receiving second sensor data; and
segmenting the second sensor data based at least in part on the sensor data segmentation model.
14. The method of claim 9 , wherein:
segmenting the image data comprises applying an operation to the first sensor data in accordance with a first stride along a first direction and a second stride, different from the first stride, along a second direction, and
training the sensor data segmentation model comprises computing a loss, the loss based at least in part on an output of the sensor data segmentation model and the training dataset, the loss comprising one or more of a cross-entropy softmax loss, a focal loss, or a logistic regression loss.
15. The method of claim 9 , further comprising:
identifying discontinuities in portions of the first sensor data; and
disregarding the portions of the first sensor data associated with the discontinuities.
16. The method of claim 9 , further comprising:
receiving second sensor data from a sensor; and
segmenting, as segmented second sensor data, the second sensor data received from the sensor using the sensor data segmentation model.
17. The method of claim 16 , further comprising generating a trajectory for an autonomous vehicle based at least in part on the segmented second sensor data.
18. One or more non-transitory computer-readable storage media having computer-executable instructions stored thereupon which, when executed by one or more processors, cause the one or more processors to perform operations comprising:
receiving image data;
storing, based at least in part on the image data, a segmented image;
receiving sensor data generated by a sensor comprising at least one of a light detection and ranging (LIDAR) sensor, a radio detection and ranging (RADAR) sensor, or a sound navigation and ranging (SONAR) sensor;
associating the sensor data with the segmented image to create a training dataset; and
training a sensor data segmentation model using the training dataset such that, upon receiving input comprising at least one of additional LIDAR sensor data, additional RADAR sensor data, or additional SONAR sensor data, the sensor data segmentation model is configured to output at least one of segmented additional LIDAR data, segmented additional RADAR sensor data, or segmented additional SONAR sensor data.
19. The non-transitory computer-readable storage media of claim 18 , wherein the sensor data is generated by a light detection and ranging (LIDAR) sensor, and
wherein the operations further comprise transmitting the sensor data segmentation model to a vehicle configured to generate a trajectory based at least in part on an additional output of the sensor data segmentation model.
20. The non-transitory computer-readable storage media of claim 18 , wherein:
a first portion of the image data is captured at a first image time;
a second portion of image data is captured at a second image time;
a first portion of the sensor data is associated with a first sensor time;
a second portion of the sensor data is associated with a second sensor time; and
the operations further comprise:
determining a first time difference between the first image time and the first sensor time;
determining a second time difference between the second image time and the first sensor time;
determining that the first time difference is less than the second time difference; and
associating the first portion of the sensor data with the first portion of the image data based at least in part on the first time difference being less than the second time difference.