Abstract: An apparatus, method and computer program product are provided for predicting feature space decay using variational auto-encoder networks. Methods may include: receiving a first image of a road segment including a feature disposed along the road segment; applying a loss function to the feature of the first image; generating a revised image, where the revised image includes a weathered iteration of the feature; generating a predicted image using interpolation between the image and the revised image of a partially weathered iteration of the feature; receiving a user image, where the user image is received from a vehicle traveling along the road segment; correlating a feature in the user image to the partially weathered iteration of the feature in the predicted image; and establishing that the feature in the user image is the feature disposed along the road segment.

Abstract: Methods described herein relate to creating a visual map of an environment free of scene clutter. Methods may include: receiving sensor data from at least one image sensor, where the sensor data is representative of a plurality of images, each image representative of a scene at a scene location; processing each image using semantic scene segmentation to identify segments of the image of the scene; classifying the segments of each of the images of the scene into one of static elements or dynamic elements; generating a decluttered image of the scene, where the decluttered image includes only elements classified as static elements; providing for storage of the decluttered image of the scene in a database; and identifying a location of a device as the scene location in response to sensor data from the device corresponding to the decluttered image of the scene.

Abstract: An approach is provided for providing a lower-power perception architecture. The approach involves, for example, determining that a device is equipped with a first perception system and a second perception system. The second perception system operates in a lower-power consumption mode than the first perception system to process image data for image recognition. The approach also involves determining a battery level of the device. The approach further involves switching from the first perception system to the second perception system based on determining that the battery level is below a threshold battery level.

Abstract: A method, apparatus and computer program product are provided for warping a perspective image into the ground plane using a homography transformation to estimate a bird's eye view in real time. Methods may include: receiving first sensor data from a first vehicle traveling along a road segment in an environment, where the first sensor data includes perspective image data of the environment, and where the first sensor data includes a location and a heading; retrieving a satellite image associated with the location and heading; applying a deep neural network to regress a bird's eye view image from the perspective image data; applying a Generative Adversarial Network (GAN) to the regressed bird's eye view image using the satellite image as a target of the GAN to obtain a stabilized bird's eye view image; and deriving values of a homography matrix between the sensor data and the established bird's eye view image.

Abstract: Methods described herein relate to identifying critical parameters of a localization framework that may require tuning for accuracy of the localization framework based on the data source. Methods may include: receiving a data set from a data source; comparing the data set against data stored in a database; identifying, based on the comparison, a subset of parameters of a localization framework for the data set from among a plurality of parameters of the localization framework; providing for tuning of the subset of parameters of the localization framework for the data set to generate a tuned subset of parameters; processing the data set from the data source using the localization framework including the tuned subset of parameters; and receiving an indication of a location from the localization framework of the data set within a mapped region.

Abstract: An approach is provided for aligning one or more drive segments based on a consensus set of user defined inputs. The approach involves, for example, retrieving manual drive alignment data collected from a plurality of human users, wherein the manual drive alignment data indicates one or more regions of at least two drive data segments selected by the human users to align the at least two drive segments. The approach also involves processing the manual drive alignment data to determine a set of common drive alignment locations of the one or more regions. The approach further involves processing a plurality of subsequent drive segments to automatically align the plurality of subsequent drive segments based on the set of common drive alignment locations.

Abstract: Methods described herein relate to reducing the computational intensity of vision-based localization. Methods may include: receiving sensor data from a vehicle traveling along a road; identifying one or more features of the environment from the sensor data; classifying the one or more identified features into one or more of a plurality of semantic classifications for the features; identifying map image data based on an identified location of the vehicle; identifying one or more features in the map image data; comparing one or more identified features of a first semantic classification with one or more features of the map image data of the first semantic classification; and registering a localized location of the vehicle within the environment based, at least in part, on the one or more identified features of the first semantic classification corresponding to the one or more features of the map image data of the first semantic classification.

Abstract: A method, apparatus and computer program product are provided for warping a perspective image into the ground plane using a homography transformation to estimate a bird's eye view in real time. Methods may include: receiving first sensor data from a first vehicle traveling along a road segment in an environment, where the first sensor data includes perspective image data of the environment, and where the first sensor data includes a location and a heading; retrieving a satellite image associated with the location and heading; applying a deep neural network to regress a bird's eye view image from the perspective image data; applying a Generative Adversarial Network (GAN) to the regressed bird's eye view image using the satellite image as a target of the GAN to obtain a stabilized bird's eye view image; and deriving values of a homography matrix between the sensor data and the established bird's eye view image.

Abstract: Methods, systems, apparatuses, and computer program products are provided that are configured to perform localization of position data, specifically using a trained localization neural network. In the context of an apparatus, the apparatus is caused to receive observed feature representation data. The apparatus is further configured to transform the observed feature representation data into standardized feature representation data utilizing a trained localization neural network. The apparatus is further configured to compare the standardized feature representation data and the map feature representation data and identify local position data based on the comparison.

Abstract: Described herein are methods of generating learning data to facilitate de-biasing the labeled location of an object of interest within an image. Methods may include: receiving sensor data, where the sensor data is a first image; determining reference corner locations of an object in the first image using image processing; generating observed corner locations of the object in the first image from the determined reference corner locations; generating a bias transformation based, at least in part, on a difference between the reference corner locations and the observed corner locations of the object in the first image; receiving sensor data from another image sensor of a second image; receiving observed corner locations of an object in the second image from a user; and applying the bias transformation to the observed corner locations of the object in the second image to generate de-biased corners for the object in the second image.

Abstract: A method is provided for generating a representation of an environment. Methods may include: determining location information of a vehicle including a road segment and a direction of travel; identifying features of the road segment based on sensor data from sensors carried by the vehicle; projecting the features of the road segment onto a ground plane of the road segment; defining bins across a width of the road segment; laterally positioning the defined bins relative to a determination of positions of the features of the road segment; consolidating detected features of each bin to define features of the road segment; and guiding an autonomous vehicle along the road segment based, at least in part, on the consolidated detected features of the road segment.

Abstract: A method is provided for dynamically throttling processing and memory consumption of sensors based on context. Methods may include: receiving first sensor data from a first sensor, where the first sensor data includes data associated with an environment of the sensor; applying an auto-encoder framework to the first sensor data to establish a data difference score between frames of the first sensor data, where a relatively high data difference score corresponds to substantial differences between frames of the first sensor data, and wherein a relatively low data difference score corresponds to insubstantial differences between frames of the first sensor data; reducing a frame rate of data capture of the first sensor in response to the data difference score between frames of the first sensor data being relatively low; capturing first sensor data from the first sensor at the reduced frame rate; and providing for storage of the first sensor data.

Abstract: A method, apparatus and computer program product are provided for enforcing a differential policy for autonomous vehicles along a road. In the context of a method, the may include: determining location information of a vehicle including a road segment and a direction of travel; identifying capabilities of sensors of the vehicle; determining an autonomous or semi-autonomous policy for the road segment specific to the vehicle in response to identifying the capabilities of the sensors of the vehicle; providing a first set of instructions for autonomous control in response to the capabilities of the sensors of the vehicle satisfying a first level of capability; and providing a second set of instructions for autonomous control in response to the capabilities of the sensors of the vehicle satisfying a second level of capability.

Abstract: Described herein are methods of generating learning data to facilitate de-biasing the labeled location of an object of interest within an image. Methods may include: receiving sensor data, where the sensor data is a first image; determining reference corner locations of an object in the first image using image processing; generating observed corner locations of the object in the first image from the determined reference corner locations; generating a bias transformation based, at least in part, on a difference between the reference corner locations and the observed corner locations of the object in the first image; receiving sensor data from another image sensor of a second image; receiving observed corner locations of an object in the second image from a user; and applying the bias transformation to the observed corner locations of the object in the second image to generate de-biased corners for the object in the second image.

Abstract: A method is provided for comparing relevant information between observations. Methods may include: receiving first sensor data from a first sensor type; receiving second sensor data from a second sensor of a second sensor type, different from the first sensor type, where the first and second sensor data includes data associated with an environment of the sensors; generating a first binary bitmask of the first sensor data; generating a second binary bitmask of the second sensor data; applying a mutual information score function using the first binary bitmask and the second binary bitmask as inputs; generating a mutual information score from the mutual information score function, wherein the mutual information score represents a degree of similarity between the sensor data; and establishing a reliability of the at least one of the first sensor or the second sensor.

Abstract: Methods described herein relate to reducing the computational intensity of vision-based localization. Methods may include: receiving sensor data from a vehicle traveling along a road; identifying one or more features of the environment from the sensor data; classifying the one or more identified features into one or more of a plurality of semantic classifications for the features; identifying map image data based on an identified location of the vehicle; identifying one or more features in the map image data; comparing one or more identified features of a first semantic classification with one or more features of the map image data of the first semantic classification; and registering a localized location of the vehicle within the environment based, at least in part, on the one or more identified features of the first semantic classification corresponding to the one or more features of the map image data of the first semantic classification.

Abstract: Methods, systems, apparatuses, and computer program products are provided that are configured to perform localization of position data, specifically using a trained localization neural network. In the context of an apparatus, the apparatus is caused to receive observed feature representation data. The apparatus is further configured to transform the observed feature representation data into standardized feature representation data utilizing a trained localization neural network. The apparatus is further configured to compare the standardized feature representation data and the map feature representation data and identify local position data based on the comparison.

Abstract: A method, apparatus and computer program product are provided for establishing the reliability of crowd sourced data based on the source of the data and the region in which the data is gathered. Methods may include: receiving map data for a network of roads in a geographic area; receiving location accuracy data from different regions within the geographic area, where the location accuracy data for each region may be indicative of an accuracy with which location can be established in the respective region; receiving a plurality of probe data points; determining, for each probe data point, the location accuracy data associated with the location information associated with the respective probe apparatus; determining, for each probe apparatus, a reliability of the one or more sensors; and updating a map database based on at least one probe data point.

Abstract: A method, apparatus and computer program product are provided to incentivize travel along specific road segments. Methods may include: receiving map data for a network of roads, where the map data includes data representing road segments within the network of roads; identifying road segments within the network of roads for which updated map data is needed; receiving an indication of generation of a route from an origin to a destination, where an alternative route is available that traverses at least one road segment for which updated map data is needed; generating an incentive for a user associated with the route that traverses at least one road segment for which updated map data is needed; providing an indication of the incentive to the user; and receiving probe data from a vehicle associated with the user having traversed the at least one road segment for which updated map data is needed.

Abstract: An approach is provided for feature triangulation from images. The approach includes retrieving the plurality of images, wherein the plurality of images is captured by a sensor of a vehicle during a drive; determining a vehicle trajectory of the vehicle during the drive; selecting a first image and a second image from the plurality of images, wherein the first image and second image are arranged in reverse time order based on respective image capture times determined using the vehicle trajectory; after detecting the feature in the first image, processing the second image to detect the feature and to associate the feature as detected in the second image with the feature previously detected in the first image; and processing the detected feature in the first image and the second image to triangulate the location of the feature.