classification

Bonnetal Classification

This readme contains simple usage explanations of how the inference works with the C++ interface, both with PyTorch and TensorRT. Before you start with the inference, make sure you've read the readme in the training part, so that you know how to generate an inference model.

Example of a PhD student being ignored by an algorithm (and why full image classification is a bad idea). This is from a ResNet50 trained on ImageNet (person is not a class, and neither is PhD student 😛).

---

Build

Standalone

$ cd bonnetal/deploy
$ git clone https://github.com/ros/catkin.git src/catkin
$ CMAKE_PREFIX_PATH=""
$ catkin init
$ catkin build bonnetal_classification_standalone

If you get an error, and your system default is python3 (3.5/3.6) you may need to change the default to avoid breaking the catkin-pkg package:

$ sudo update-alternatives --install /usr/bin/python python /usr/bin/python2.7 1
$ sudp update-alternatives --install /usr/bin/python python /usr/bin/python3.6 2

ROS

Extending an existing catkin workspace (ROS installed and workspace sourced):

$ cd bonnetal/deploy
$ catkin init
$ catkin build
$ source devel/setup.bash

---

Perform Inference

Standalone

• Infer images:

  # use --verbose or -v to get verbose mode
  $ ./devel/lib/bonnetal_classification_standalone/infer_img -h # help
  $ ./devel/lib/bonnetal_classification_standalone/infer_img -p /path/to/save/inference/ready -i /path/to/images # can be multiple images
  $ ./devel/lib/bonnetal_classification_standalone/infer_img -p /path/to/save/inference/ready -i /path/to/images -b pytorch
  $ ./devel/lib/bonnetal_classification_standalone/infer_img -p /path/to/save/inference/ready -i /path/to/images -b tensorrt

• Infer video/webcam:

  # use --verbose or -v to get verbose mode
  $ ./devel/lib/bonnetal_classification_standalone/infer_video -h # help
  $ ./devel/lib/bonnetal_classification_standalone/infer_video -p /path/to/save/inference/ready --video /path/to/video
  $ ./devel/lib/bonnetal_classification_standalone/infer_video -p /path/to/save/inference/ready --video /path/to/video -b pytorch
  $ ./devel/lib/bonnetal_classification_standalone/infer_video -p /path/to/save/inference/ready --video /path/to/video -b tensorrt

• Infer Webcam instead of video:

    # use --verbose or -v to get verbose mode
    $ ./devel/lib/bonnetal_classification_standalone/infer_video -p /path/to/pretrained/ # simply drop the video and it will look for /dev/video0

ROS

• Example with webcam contains all necessary stuff (see the config file inside the _ros package):

  $ roslaunch bonnetal_classification_ros bonnetal_classification_webcam_sample.launch