Methods and apparatus for natural language-based safety case discovery to train a machine learning model for a driving system
AssigneePlusAI, Inc.
InventorsHao Zheng, Yang Yang
Life of this patent
Estimating the term…
Abstract
A safety case discovery system includes a scenario framework and safety protocols for edge cases. The safety case discovery system receives sensor data generated by at least one sensor during operation of a vehicle and stores the sensor data in a data warehouse. The data warehouse can be queried based on a predefined scenario description to produce a subset of records which are ranked based on a relevancy of the records to the predefined scenario description. The safety case discovery system deduplicates the ranked results to produce edge cases and updates the safety framework. The safety case discovery system can train a machine learning model vehicle based on the edge cases.
Claims20
An apparatus, comprising:
- a processor;
- and
- a memory operably coupled to the processor, the memory storing instructions to cause the processor to:
- receive sensor data generated via at least one sensor during operation of a vehicle;
- identify a plurality of features associated with the sensor data;
- generate a multidimensional representation of the sensor data based on the plurality of features;
- query a queryable search engine index based on the multidimensional representation of the sensor data, to produce a query result including a subset of records from a plurality of records of the queryable search engine index;
- rank each record from the subset of records based on a relevancy of that record to the multidimensional representation of the sensor data, to produce a ranked set of records;
- identify at least one edge case based on the ranked set of records;
- and
- train a machine learning model associated with the vehicle based on the at least one edge case and training data, to produce a trained machine learning model.
A non-transitory, processor-readable medium storing instructions to cause a processor to:
- receive sensor data generated by at least one sensor located at an autonomous vehicle during operation of the autonomous vehicle;
- identify a plurality of features associated with the sensor data;
- generate a multidimensional representation of the sensor data based on the plurality of features;
- query a queryable search engine index based on the multidimensional representation of the sensor data, to produce a query result including a subset of records from a plurality of records of the queryable search engine index;
- rank each record from the subset of records based on a relevancy of that record to the multidimensional representation of the sensor data, to produce a ranked set of records;
- receive a signal representing a user selection of at least one edge case from the ranked set of records for inclusion in a safety framework;
- and
- cause an update to the safety framework such that the safety framework includes the at least one edge case.
A method comprising:
- receiving, via a processor operatively coupled to a memory, sensor data generated by at least one sensor located at a vehicle during operation of the vehicle;
- identifying a plurality of features associated with the sensor data;
- generating a multidimensional representation of the sensor data based on the plurality of features;
- querying a queryable search engine index based on the multidimensional representation of the sensor data, to produce a query result including a subset of records from a plurality of records of the queryable search engine index;
- ranking each record from the subset of records based on a relevancy of that record to the multidimensional representation of the sensor data, to produce a ranked set of records;
- receiving a signal representing a user selection of at least one edge case from the ranked set of records for inclusion in a safety framework;
- and
- causing an update to the safety framework such that the safety framework includes the at least one edge case.
Dependencies are inferred from the claim text.