Object tracking using semantic attributes
AssigneeGM Cruise Holdings LLC
InventorHariprasad Govardhanam
Life of this patent
Estimating the term…
Abstract
The disclosed technology provides solutions for improving object tracking. In some aspects, a process of the disclosed technology can include steps for receiving Light Detection and Ranging (LiDAR) point cloud data corresponding with an object, providing the LiDAR point cloud data to an autoencoder neural network to generate a set of feature embeddings, and tracking the object based on the feature embeddings. Systems and machine-readable media are also provided.
Claims20
An apparatus for performing object tracking, comprising:
- at least one memory;
- and
- at least one processor coupled to the at least one memory, the at least one processor configured to:
- receive a first set of Light Detection and Ranging (LiDAR) point cloud data corresponding with an object;
- provide the first set of LiDAR point cloud data to an autoencoder neural network to generate a first set of feature embeddings, the first set of feature embeddings comprising one or more semantic attributes associated with the object;
- and
- track a trajectory of the object based on the first set of feature embeddings and one or more physical attributes associated with the object.
A computer-implemented method for tracking an object, comprising:
- receiving a first set of Light Detection and Ranging (LiDAR) point cloud data corresponding with the object;
- providing the first set of LiDAR point cloud data to an autoencoder neural network to generate a first set of feature embeddings, the first set of feature embeddings comprising one or more semantic attributes associated with the object;
- and
- tracking a trajectory of the object based on the first set of feature embeddings and one or more physical attributes associated with the object.
A non-transitory computer-readable storage medium comprising at least one instruction for causing a computer or processor to:
- receive a first set of Light Detection and Ranging (LiDAR) point cloud data corresponding with an object;
- provide the first set of LiDAR point cloud data to an autoencoder neural network to generate a first set of feature embeddings, the first set of feature embeddings comprising one or more semantic attributes associated with the object;
- and
- track a trajectory of the object based on the first set of feature embeddings and one or more physical attributes associated with the object.
Dependencies are inferred from the claim text.