R2Inference - TensorRT Examples
< R2Inference | Examples
Make sure you also check R2Inference's companion project: GstInference |
R2Inference |
---|
Introduction |
Getting started |
Supported backends |
Examples |
Model Zoo |
Contact us |
|
Contents
R2Inference provides several examples to highlight its usage with different architectures and frameworks. The models used for these examples can be downloaded from our Model Zoo.
Preparation
To test the TensorRT example you will need a TensorRT C++ API compatible graph file. You can obtain the model from the Model Zoo.
Yolo v2
This example is located in r2inference/examples/r2i/tensorrt
.
To use this example run:
./yolov2 -i [JPG input Image] -m [TRT Model]
For example, evaluating this image:
Error creating thumbnail: Unable to save thumbnail to destination
Should produce the following output:
./yolov2 -i Plane_example_input.jpg -m graph_yolov2.trt
Loading Model: graph_yolov2.trt
Setting model to engine
Loading image: Plane_example_input.jpg
Configuring frame
Starting engine
Predicting...
prob: 12.11 label: 0
x: 148 y: 68 width: 130 height: 107
prob: 12.30 label: 0
x: 71 y: 51 width: 275 height: 138
prob: 8.96 label: 0
x: 151 y: 67 width: 160 height: 108
prob: 9.24 label: 0
x: 103 y: 51 width: 266 height: 136
prob: 14.02 label: 0
x: 34 y: 75 width: 296 height: 126
prob: 12.29 label: 0
x: 149 y: 70 width: 127 height: 129
prob: 13.00 label: 0
x: 70 y: 78 width: 279 height: 118
prob: 8.39 label: 0
x: 152 y: 80 width: 158 height: 109
prob: 9.54 label: 0
x: 99 y: 79 width: 280 height: 118
Stopping engine
According to the VOC labels 0 corresponds to an 'aeroplane'.