# US 12,600,384 B2 · Model hyperparameter adjustment using vehicle driving context classification

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,600,384 B2
- Document ID: US12600384B2
- Status: Granted
- Kind: B2
- Application number: 17/655,837
- Filed: 2022-03-22
- Published: 2023-09-28
- Granted: 2026-04-14
- Earliest claimed: 2022-03-22
- Assignees: QUALCOMM Incorporated
- Applicants: QUALCOMM Incorporated
- Inventors: Ahmed Kamel Sadek; Pranav Desai; Stephen Marc Chaves; Paul Daniel Martin
- Web page: https://patentgrep.com/patent/US12600384B2

## CPC

- B60W 60/0027: {using trajectory prediction for other traffic participants}
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- B60W 50/0097: {Predicting future conditions}
- G05B 13/027: {using neural networks only}
- G06N 3/008: based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
- G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn
- B60W 2050/0028: {Mathematical models, e.g. for simulation}
- B60W 2420/403: Image sensing, e.g. optical camera
- B60W 2552/05: Type of road, e.g. motorways, local streets, paved or unpaved roads
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/0499: Feedforward networks
- G06N 3/092: Reinforcement learning
- G06N 3/098: Distributed learning, e.g. federated learning
- B60W: CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)

## Abstract

In some aspects, a device of a vehicle may obtain information relating to an environment in which the vehicle is located. The device may determine using a machine learning model, a driving context of the vehicle based at least in part on the information relating to the environment, and a set of hyperparameters for a model, that is used to determine a driving behavior for the vehicle, based at least in part on the driving context. The device may determine, using the model configured with the set of hyperparameters, the driving behavior for the vehicle. The device may cause autonomous operation of the vehicle in accordance with the driving behavior. Numerous other aspects are described.

## Claims

### Claim 1

A method, comprising:
obtaining, by a device of a vehicle, information relating to an environment in which the vehicle is located;
determining, by the device, using a machine learning model:
a current driving context of the vehicle based at least in part on the information relating to the environment, and
a plurality of hyperparameters, comprising a tree width, a tree depth, a time length for a step in a tree, and a reward function, wherein each hyperparameter, of the plurality of hyperparameters, is determined based at least in part on the current driving context, for a model, wherein the model is used to determine an autonomous driving behavior for the vehicle;
configuring the model with the plurality of hyperparameters based at least in part on determining the plurality of hyperparameters;
determining, by the device and using the model configured with the plurality of hyperparameters, the autonomous driving behavior, wherein the autonomous driving behavior is determined based at least in part on the reward function for the model; and
causing, by the device, autonomous operation of the vehicle in accordance with the autonomous driving behavior.

### Claim 2 (depends on 1)

The method of claim 1,
wherein the model is based at least in part on a heuristic search algorithm.

### Claim 3 (depends on 1)

The method of claim 1,
wherein the model is based at least in part on a Monte Carlo tree search.

### Claim 4 (depends on 1)

The method of claim 1,
wherein the machine learning model is a neural network trained to determine the plurality of hyperparameters for the model based at least in part on the current driving context.

### Claim 5 (depends on 1)

The method of claim 1,
wherein the current driving context of the vehicle is an intersection context, a highway context, a local road context, or a parking lot context.

### Claim 6 (depends on 1)

The method of claim 1,
wherein the information relating to the environment is obtained from one or more sensors of the vehicle.

### Claim 7 (depends on 6)

The method of claim 6,
wherein the one or more sensors include one or more of: a light detection and ranging system or a camera.

### Claim 8 (depends on 1)

The method of claim 1,
wherein the autonomous driving behavior is determined to maximize the reward function for the model.

### Claim 9 (depends on 1)

The method of claim 1, further comprising:
determining one or more intention predictions relating to one or more objects in the environment,
wherein at least one of the current driving context or the plurality of hyperparameters is determined further based at least in part on the one or more intention predictions.

### Claim 10 (depends on 1)

The method of claim 1, further comprising:
receiving, from another vehicle, an indication of the plurality of hyperparameters, wherein the determining the plurality of hyperparameters based at least in part on the indication.

### Claim 11 (depends on 10)

The method of claim 10, wherein the indication further comprises a feedback value, wherein the feedback value indicates an accuracy associated with the autonomous driving behavior.

### Claim 12 (depends on 11)

The method of claim 11, wherein the feedback value comprises a loss function value or value for the reward function.

### Claim 13 (depends on 11)

The method of claim 11, wherein configuring the model further comprises:
configuring the model with the feedback value.

### Claim 14

A system for autonomous driving of a vehicle, comprising:
one or more memories; and
one or more processors, coupled to the one or more memories, configured to:
obtain information relating to an environment in which the vehicle is located;
determine, using a machine learning model:
a current driving context of the vehicle based at least in part on the information relating to the environment, and
a plurality of hyperparameters, comprising a tree width, a tree depth, a time length for a step in a tree, and a reward function, wherein each hyperparameter, of the plurality of hyperparameters, is determined based at least in part on the current driving context, for a model, wherein the model is used to determine an autonomous driving behavior for the vehicle;
configure the model with the plurality of hyperparameters based at least in part on determining the plurality of hyperparameters;
determine, using the model configured with the plurality of hyperparameters, the autonomous driving behavior, wherein the autonomous driving behavior is determined based at least in part on the reward function for the model; and
cause autonomous operation of the vehicle in accordance with the autonomous driving behavior.

### Claim 15 (depends on 14)

The system of claim 14,
wherein the model is based at least in part on a heuristic search algorithm.

### Claim 16 (depends on 14)

The system of claim 14,
wherein the model is based at least in part on a Monte Carlo tree search.

### Claim 17 (depends on 14)

The system of claim 14,
wherein the machine learning model is a neural network trained to determine the plurality of hyperparameters for the model based at least in part on the current driving context.

### Claim 18 (depends on 14)

The system of claim 14,
wherein the current driving context of the vehicle is an intersection context, a highway context, a local road context, or a parking lot context.

### Claim 19 (depends on 14)

The system of claim 14,
wherein the information relating to the environment is obtained from one or more sensors of the vehicle.

### Claim 20 (depends on 19)

The system of claim 19,
wherein the one or more sensors include one or more of: a light detection and ranging system or a camera.

### Claim 21 (depends on 14)

The system of claim 14,
wherein the autonomous driving behavior is determined to maximize the reward function for the model.

### Claim 22 (depends on 14)

The system of claim 14,
wherein the one or more processors are further configured to:
determine one or more intention predictions relating to one or more objects in the environment,
wherein at least one of the current driving context or the plurality of hyperparameters is determined further based at least in part on the one or more intention predictions.

### Claim 23

A non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising:
one or more instructions that, when executed by one or more processors of a device, cause the device to:
obtain information relating to an environment in which a vehicle is located;
determine, using a machine learning model:
a current driving context of the vehicle based at least in part on the information relating to the environment, and
a plurality of hyperparameters, comprising a tree width, a tree depth, a time length for a step in a tree, and a reward function, wherein each hyperparameter, of the plurality of hyperparameters, is determined based at least in part on the current driving context, for a model, wherein the model is used to determine an autonomous driving behavior for the vehicle;
configure the model with the plurality of hyperparameters based at least in part on determining the plurality of hyperparameters;
determine, using the model configured with the plurality of hyperparameters, the autonomous driving behavior, wherein the autonomous driving behavior is determined based at least in part on the reward function for the model; and
cause autonomous operation of the vehicle in accordance with the autonomous driving behavior.

### Claim 24 (depends on 23)

The non-transitory computer-readable medium of claim 23,
wherein the current driving context of the vehicle is an intersection context, a highway context, a local road context, or a parking lot context.

### Claim 25 (depends on 23)

The non-transitory computer-readable medium of claim 23,
wherein the one or more instructions further cause the device to:
determine one or more intention predictions relating to one or more objects in the environment,
wherein at least one of the current driving context or the plurality of hyperparameters is determined further based at least in part on the one or more intention predictions.

### Claim 26

An apparatus, comprising:
means for obtaining information relating to an environment in which a vehicle is located;
means for determining, using a machine learning model,
a current driving context of the vehicle based at least in part on the information relating to the environment, and
a plurality of hyperparameters, comprising a tree width, a tree depth, a time length for a step in a tree, and a reward function, wherein each hyperparameter, of the plurality of hyperparameters, is determined based at least in part on the current driving context, for a model, wherein the model is used to determine an autonomous driving behavior for the vehicle;
means for configuring the model with the plurality of hyperparameters based at least in part on determining the plurality of hyperparameters;
means for determining, by the apparatus and using the model configured with the plurality of hyperparameters, the autonomous driving behavior, wherein the autonomous driving behavior is determined based at least in part on the reward function for the model; and
means for causing, by the apparatus, autonomous operation of the vehicle in accordance with the autonomous driving behavior.

### Claim 27 (depends on 26)

The apparatus of claim 26, further comprising:
means for determining one or more intention predictions relating to one or more objects in the environment,
wherein the plurality of hyperparameters is determined further based at least in part on the one or more intention predictions.

### Claim 28 (depends on 26)

The apparatus of claim 26,
wherein the information relating to the environment is obtained from one or more sensors of the vehicle.

### Claim 29 (depends on 28)

The apparatus of claim 28,
wherein the one or more sensors include one or more of: a light detection and ranging system or a camera.

### Claim 30 (depends on 26)

The apparatus of claim 26,
wherein the autonomous driving behavior is determined to maximize the reward function for the model.