How to change options of imagenet-console

Hi.

I want to change the options of imagenet-console like prototxt, batch size etc.

I tried to change prototxt but it failed.
[b]
nvidia@tegra-ubuntu:~/jetson-inference/build/aarch64/bin$ ./imagenet-console orange_0.jpg out1.jpg --prototxt=networks/alexnet.prototxt --model=networks/bvlc_alexnet.caffemodel
imagenet-console
args (5): 0 [./imagenet-console] 1 [orange_0.jpg] 2 [out1.jpg] 3 [–prototxt=networks/alexnet.prototxt] 4 [–model=networks/bvlc_alexnet.caffemodel]

imageNet – failed to initialize.
imagenet-console: failed to initialize imageNet[/b]

And I want to classify many images at once.

Please teach me. Thank you.

Hi,

  1. For correct command, please check here:
    https://github.com/dusty-nv/jetson-inference#loading-custom-models-on-jetson

  2. For multiple image input, please check here:
    https://github.com/dusty-nv/jetson-inference/blob/master/imageNet.cpp#L286

Thanks.

Hi.
The first problem solved! Thank you.

But I can’t understand the second reply.

I saw the code you linked. but I can’t understand what you said, sorry.

What should I write into command line at the location of input path and output path for classifying multiple input at once.

For multiple input, should I modify the code?
Is there any solution not to modify the code?

Hi,

You need to modify the code.

Currently, our sample passes data buffer with size 1x3xheightxwidth, which only has one image per execution.
If you want to use multiple images at the same time, you need to generate a data buffer with size kx3xheightxwidth. Then pass it into TensorRT.

Thanks.