NVIDIA DeepStream SDK API Reference

6.4 Release
gstnvinfer Namespace Reference

Data Structures

class  DsNvInferImpl
 
class  LockGMutex
 C++ helper class written on top of GMutex/GCond. More...
 
struct  ModelStatus
 Holds runtime model update status along with the error message if any. More...
 

Enumerations

enum  ModelLoadType {
  MODEL_LOAD_FROM_ENGINE,
  MODEL_LOAD_FROM_CONFIG,
  MODEL_LOAD_STOP
}
 Enum for type of model update required. More...
 

Enumeration Type Documentation

◆ ModelLoadType

Enum for type of model update required.

Enumerator
MODEL_LOAD_FROM_ENGINE 

Load a new model by just replacing the model engine assuming no network architecture changes.

MODEL_LOAD_FROM_CONFIG 

Load a new model with other configuration changes.

This option will only update the NvDsInferContext, any filtering/post-processing/pre-processing done in gst-nvinfer will not be updated. An important requirement is that the network input layer resolution should not changes.

MODEL_LOAD_STOP 

Request the model load thread to stop.

Definition at line 181 of file gstnvinfer_impl.h.