Note: SW Release Applicability: This tutorial is applicable to modules in NVIDIA DRIVE Software releases.

Initialization

DriveNet module creates a DriveNet handle ( dwDriveNetHandle_t ) that is owned by the application and must be released by the application when it's not needed anymore.

The following dwDriveNetParams are required to be defined at initialization:

uint32_t maxProposalsPerClass

Maximum number of proposals (i.e. detections before clustering) per class. This value can only be set at initialization.

uint32_t maxClustersPerClass

Maximum number of clusters per class. This value can only be set at initialization.

dwPrecision networkPrecision

This parameter determines at which precision DriveNet must be loaded. This parameter can affect the speed of the network greatly.

bool hasObjectDepth

This parameter determines if we wish enable depth estimation for objects the detected.

DriveNet module can be initialized with the following function:

dwStatus dwDriveNet_initialize(dwDriveNetHandle_t* drivenetHandle,
                               const dwDriveNetParams* drivenetParams,
                               dwContextHandle_t ctx);

DriveNet provides at initialization as outputs the number of object classes that DriveNet can detect and the list of object classes that DriveNet can detect. The object classes are returned as dwObjectClass. Upon desire, the classes can be obtained via dwDriveNet_getAvailableClasses().

DriveNet comes with meta data which contains preprocessing configuration for the network. Object Detector module automatically reads this meta data information and configures Data Conditioner accordingly. Object Detector Workflow shows in more details how to use DriveNet after initialization.

For more details see DriveNet Sample