ROS IndexNo version for distro humble. Known supported distros are highlighted in the buttons above.
No version for distro jazzy. Known supported distros are highlighted in the buttons above.
No version for distro rolling. Known supported distros are highlighted in the buttons above.
Package Summary
| Tags | No category tags. |
| Version | 0.0.0 |
| License | Apache-2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | ROS package for OpenTera WebRTC signaling server. |
| Checkout URI | https://github.com/introlab/opentera-webrtc-ros.git |
| VCS Type | git |
| VCS Version | ros2 |
| Last Updated | 2025-02-14 |
| Dev Status | UNKNOWN |
| CI status | No Continuous Integration |
| Released | UNRELEASED |
| Tags | robotics webrtc ros teleoperation opentera |
| Contributing |
Help Wanted (0)
Good First Issues (0) Pull Requests to Review (0) |
Package Description
The face_cropping package
Additional Links
No additional links.
Maintainers
- Dominic Létourneau
Authors
No additional authors.
face_cropping
Face Detection Models
- haarcascade_frontalface_default.xml Haar feature-based cascade classifiers to detect faces. The model is from the OpenCV repository and the license is at the beginning of the file.
- lbpcascade_frontalface_improved.xml LBP feature-based cascade classifiers to detect faces. The model is from the OpenCV repository and the license is at the beginning of the file.
- small_yunet_.pt The models are derived from YuNet, but it is faster since it is optimized for big faces. Their backbones are trained on ImageNet using knowledge distillation. The face detectors are trained on WIDER FACE without the small faces. The training code is available in the dnn_training folder. SimOTA is used for the face assignation. The SimOTA code is from (MMDetection)[https://github.com/open-mmlab/mmdetection].
LibTorch
LibTorch and torchvision are required to use small_yunet_* models. LibTorch and torchvision are automatically downloaded on AMD64 computers. On Jetson computer, you need to install PyTorch built by NVIDIA and compile torchvision for C++.
To use the GPU, you need to set the option FACE_CROPPER_USE_CUDA to ON.
Resource usage
The CPU and memory usage are measured on an AMD RYZEN 7 3700X at 30 Hz. The APs are measured on a subset of WIDER FACE. The subset includes faces occupying at least 10% of the width and the height of the image.
| Model | LibTorch Required | CPU (%, 100% = 1 core) | Memory (MB) | AP@0.25 | AP@0.50 | AP@0.75 |
|---|---|---|---|---|---|---|
| haarcascade | No | 220.4 | 299.7 | 0.5973 | 0.5649 | 0.0314 |
| lbpcascade | No | 154.7 | 299.1 | 0.4141 | 0.4014 | 0.0705 |
| small_yunet_0.25_160 | Yes | 11.4 | 273.9 | 0.6896 | 0.4940 | 0.1298 |
| small_yunet_0.25_320 | Yes | 16.7 | 276.4 | 0.7969 | 0.7019 | 0.2669 |
| small_yunet_0.25_640 | Yes | 34.3 | 297.4 | 0.8266 | 0.7601 | 0.4237 |
| small_yunet_0.5_160 | Yes | 12.7 | 274.4 | 0.7930 | 0.6834 | 0.3183 |
| small_yunet_0.5_320 | Yes | 20.2 | 282.2 | 0.8622 | 0.8034 | 0.4840 |
| small_yunet_0.5_640 | Yes | 50.1 | 315.8 | 0.8780 | 0.8466 | 0.6018 |
face_cropping_node
Implements a ROS node that subscribes to an image stream. It detects the biggest face in the received frame and crops it out.
It can be enabled and disabled from the robot GUI by pressing the face cropping button, by default it is enabled.
Parameters
-
face_detection_model(string): The face detection model (haarcascade,lbpcascade,small_yunet_0.25_160,small_yunet_0.25_320,small_yunet_0.25_640,small_yunet_0.5_160,small_yunet_0.5_320orsmall_yunet_0.5_640). -
use_gpu_if_available(bool): Indicates whether to use the GPU or not. -
min_face_width(double): The minimum face width. -
min_face_height(double): The minimum face height. -
output_width(int): The output face image width. -
output_height(int): The output face image height. -
adjust_brightness(bool): Indicates whether to adjust the brightness or not.
Subscribed Topics
-
enable_face_croppingstd_msgs/Bool: The topic to enable or disable the node. -
input_imagesensor_msgs/Image: The input image topic.
Published Topics
-
output_imagesensor_msgs/Image: The output image topic.
CHANGELOG
No CHANGELOG found.
Wiki Tutorials
This package does not provide any links to tutorials in it's rosindex metadata.
You can check on the ROS Wiki Tutorials page for the package.
Package Dependencies
| Deps | Name |
|---|---|
| ament_cmake | |
| ament_lint_auto | |
| ament_lint_common | |
| rclcpp | |
| rclpy | |
| sensor_msgs | |
| opentera_webrtc_ros_msgs | |
| cv_bridge | |
| image_transport | |
| ament_index_cpp |
System Dependencies
No direct system dependencies.
Dependant Packages
No known dependants.
Launch files
- launch/face_cropping.launch.xml
-
- name [default: face_cropping_node]
- input_camera_topic [default: /camera1/image_raw]
- output_camera_topic [default: /face_camera/image_raw]
- face_detection_model [default: lbpcascade]
- use_gpu_if_available [default: false]
- min_face_width [default: 100.0]
- min_face_height [default: 100.0]
- output_width [default: 480]
- output_height [default: 720]
- adjust_brightness [default: true]
- launch/test_face_cropping.launch.xml
-
- model [default: lbpcascade]
- launch/face_cropping_webrtc.launch.xml
-
- name [default: face_cropping_webrtc_node]
- input_camera_topic [default: /camera1/image_raw]
- output_camera_topic [default: /face_camera/image_raw]
- face_detection_model [default: lbpcascade]
- use_gpu_if_available [default: false]
- min_face_width [default: 100.0]
- min_face_height [default: 100.0]
- output_width [default: 480]
- output_height [default: 720]
- adjust_brightness [default: true]
- launch/test_face_cropping_webrtc.launch.xml
-
- model [default: lbpcascade]
Messages
No message files found.
Services
No service files found
Plugins
No plugins found.
Recent questions tagged face_cropping at Robotics Stack Exchange
No version for distro noetic. Known supported distros are highlighted in the buttons above.
No version for distro ardent. Known supported distros are highlighted in the buttons above.
No version for distro bouncy. Known supported distros are highlighted in the buttons above.
No version for distro crystal. Known supported distros are highlighted in the buttons above.
No version for distro eloquent. Known supported distros are highlighted in the buttons above.
No version for distro dashing. Known supported distros are highlighted in the buttons above.
No version for distro galactic. Known supported distros are highlighted in the buttons above.
No version for distro foxy. Known supported distros are highlighted in the buttons above.
No version for distro iron. Known supported distros are highlighted in the buttons above.
No version for distro lunar. Known supported distros are highlighted in the buttons above.
No version for distro jade. Known supported distros are highlighted in the buttons above.
No version for distro indigo. Known supported distros are highlighted in the buttons above.
No version for distro hydro. Known supported distros are highlighted in the buttons above.
No version for distro kinetic. Known supported distros are highlighted in the buttons above.
No version for distro melodic. Known supported distros are highlighted in the buttons above.