opendr-eu · passalis · Dec 4, 2023 · Jan 17, 2023 · Feb 13, 2023 · Feb 15, 2023
@@ -133,6 +133,6 @@ jobs:
         python-version: 3.8
     - name: Test Docker
       run: |
-        docker run --name toolkit -i opendr/opendr-toolkit:cpu_v2.2.0 bash
+        docker run --name toolkit -i opendr/opendr-toolkit:cpu_v3.0.0 bash
         docker start toolkit
         docker exec -i toolkit bash -c "source bin/activate.sh && source tests/sources/tools/control/mobile_manipulation/run_ros.sh && python3 -m unittest discover -s tests/sources/tools/${{ matrix.package }}"
@@ -275,7 +275,7 @@ jobs:
     runs-on: ubuntu-20.04
     steps:
     - name: Free Disk Space
-      uses: jlumbroso/free-disk-space@v1.2.0
+      uses: jlumbroso/free-disk-space@v1.3.1
     - uses: actions/checkout@v3
       with:
         submodules: true
@@ -363,12 +363,13 @@ jobs:
         export PYTHONPATH=$OPENDR_HOME/src:$PYTHONPATH
         export ROS_DISTRO=noetic
 
-        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev python3-dev
-
+        sudo apt-get update
+        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev python3-dev python3-six ffmpeg
+
         python3 -m venv venv
         source venv/bin/activate
         python3 -m pip install --upgrade pip
-        python3 -m pip install wheel==0.38.4
+        python3 -m pip install wheel==0.38.4 six
 
         # install all tools one at a time
         while read f; do
@@ -455,12 +456,13 @@ jobs:
         export OPENDR_DEVICE=cpu
         export ROS_DISTRO=noetic
 
-        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev libeigen3-dev
+        sudo apt-get update
+        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev libeigen3-dev python3-dev python3-six ffmpeg
 
         python3 -m venv venv
         source venv/bin/activate
         python3 -m pip install --upgrade pip
-        python3 -m pip install wheel==0.38.4
+        python3 -m pip install wheel==0.38.4 six
 
         # get the name of the wheel to install based on the test being run
         package=$(sed "s/_/-/g" <<< ${{ matrix.package }})
@@ -480,6 +482,8 @@ jobs:
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-pose-estimation-*.tar.gz
           elif [ ${{ matrix.package }} == "perception/heart_anomaly_detection" ]; then
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-compressive-learning-*.tar.gz
+          elif [ ${{ matrix.package }} == "perception/gesture_recognition" ]; then
+            python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-object-detection-2d-*.tar.gz
           elif [ ${{ matrix.package }} == "perception/multimodal_human_centric" ]; then
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-compressive-learning-*.tar.gz
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-object-detection-2d-*.tar.gz
@@ -541,13 +545,13 @@ jobs:
           - perception/facial_expression_recognition
           - simulation/human_model_generation
           - control/mobile_manipulation
-          - control/single_demo_grasp
+          # - control/single_demo_grasp
           - planning/end_to_end_planning
           # - control/multi_object_search # needs CUDA Docker container
     runs-on: ubuntu-20.04
     steps:
     - name: Free Disk Space
-      uses: jlumbroso/free-disk-space@v1.2.0
+      uses: jlumbroso/free-disk-space@v1.3.1
     - name: Download artifact
       uses: actions/download-artifact@v2
       with:

@@ -279,7 +279,7 @@ jobs:
     runs-on: ubuntu-20.04
     steps:
     - name: Free Disk Space
-      uses: jlumbroso/free-disk-space@v1.2.0
+      uses: jlumbroso/free-disk-space@v1.3.1
     - uses: actions/checkout@v3
       with:
         submodules: true
@@ -369,12 +369,13 @@ jobs:
         export PYTHONPATH=$OPENDR_HOME/src:$PYTHONPATH
         export ROS_DISTRO=noetic
 
-        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev python3-dev
+        sudo apt-get update
+        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev python3-dev python3-six ffmpeg
 
         python3 -m venv venv
         source venv/bin/activate
         python3 -m pip install --upgrade pip
-        python3 -m pip install wheel==0.38.4
+        python3 -m pip install wheel==0.38.4 six
 
         # install all tools one at a time
         while read f; do
@@ -464,12 +465,13 @@ jobs:
         export PYTHONPATH=$OPENDR_HOME/src:$PYTHONPATH
         export ROS_DISTRO=noetic
 
-        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev libeigen3-dev
+        sudo apt-get update
+        sudo apt -y install python3.8-venv libfreetype6-dev git build-essential cmake python3-dev wget libopenblas-dev libsndfile1 libboost-dev libeigen3-dev python3-dev python3-six ffmpeg
 
         python3 -m venv venv
         source venv/bin/activate
         python3 -m pip install --upgrade pip
-        python3 -m pip install wheel==0.38.4
+        python3 -m pip install wheel==0.38.4 six
 
         # get the name of the wheel to install based on the test being run
         package=$(sed "s/_/-/g" <<< ${{ matrix.package }})
@@ -489,6 +491,8 @@ jobs:
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-pose-estimation-*.tar.gz
           elif [ ${{ matrix.package }} == "perception/heart_anomaly_detection" ]; then
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-compressive-learning-*.tar.gz
+           elif [ ${{ matrix.package }} == "perception/gesture_recognition" ]; then
+            python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-object-detection-2d-*.tar.gz
           elif [ ${{ matrix.package }} == "perception/multimodal_human_centric" ]; then
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-compressive-learning-*.tar.gz
             python3 -m pip install --find-links ./artifact/wheel-artifact/ ./artifact/wheel-artifact/opendr-toolkit-object-detection-2d-*.tar.gz
@@ -550,13 +554,13 @@ jobs:
           - perception/facial_expression_recognition
           - simulation/human_model_generation
           - control/mobile_manipulation
-          - control/single_demo_grasp
+          # - control/single_demo_grasp
           - planning/end_to_end_planning
           # - control/multi_object_search # needs CUDA Docker container
     runs-on: ubuntu-20.04
     steps:
     - name: Free Disk Space
-      uses: jlumbroso/free-disk-space@v1.2.0
+      uses: jlumbroso/free-disk-space@v1.3.1
     - name: Download artifact
       uses: actions/download-artifact@v2
       with:

@@ -1,5 +1,51 @@
 # OpenDR Toolkit Change Log
 
+## Version 3.0.0
+Released on Dec, 4, 2023.
+- New Features:  
+  - Binary High Resolution Learner ([#402](https://github.com/opendr-eu/opendr/pull/402))
+  - ROS2 node for EfficientLPS ([#404](https://github.com/opendr-eu/opendr/pull/404))
+  - Fall and wave detection ROS nodes ([#423](https://github.com/opendr-eu/opendr/pull/423))
+  - Continual SLAM: Adds a new SLAM tool for Continual SLAM ([#424](https://github.com/opendr-eu/opendr/pull/424))
+  - Add RGB gesture recognition ([#436](https://github.com/opendr-eu/opendr/pull/436))
+  - FSeq2-NMS ([#442](https://github.com/opendr-eu/opendr/pull/442))
+  - Intent recognition tool ([#443](https://github.com/opendr-eu/opendr/pull/443))
+  - Robotti human detection simulation demo ([#451](https://github.com/opendr-eu/opendr/pull/451))
+  - Object Detection 2D Class Filtering ([#467](https://github.com/opendr-eu/opendr/pull/467))
+  - RL-based Learner for Active Face Recognition ([#473](https://github.com/opendr-eu/opendr/pull/473))
+  - YOLOv5s Inference Demo with Optimized Weights for Agricultural Use ([#476]([#](https://github.com/opendr-eu/opendr/pull/)476))
+  - Adaptive HR Pose Estimation ([#479](https://github.com/opendr-eu/opendr/pull/479))
+- Enhancements:
+  - Wave detection demo based on pose estimation ([#394](https://github.com/opendr-eu/opendr/pull/394))
+  - Facial expression recognition demo update ([#405](https://github.com/opendr-eu/opendr/pull/405))
+  - Object detection 2d camera demos ([#408](https://github.com/opendr-eu/opendr/pull/408))
+  - High Resolution Pose Estimation webcam demo ([#409](https://github.com/opendr-eu/opendr/pull/409))
+  - ROS nodes FPS performance measurements ([#419](https://github.com/opendr-eu/opendr/pull/419))
+  - Refactoring: pythonic joins in `test_clang_format.py`/`test_cppcheck.py` ([#455](https://github.com/opendr-eu/opendr/pull/455))
+  - Test-tools improvement ([#456](https://github.com/opendr-eu/opendr/pull/456)) 
+  - Adding prompt when transcribe with Whisper ([#462](https://github.com/opendr-eu/opendr/pull/462))
+
+- Bug Fixes:
+  - Fix package creator and sources ([#390](https://github.com/opendr-eu/opendr/pull/390))
+  - Lightweight OpenPose tool fixes and improvements ([#392](https://github.com/opendr-eu/opendr/pull/392))
+  - Fall Detection - alternative infer input ([#397](https://github.com/opendr-eu/opendr/pull/397))
+  - Yolov5 training bugfix ([#401](https://github.com/opendr-eu/opendr/pull/401))
+  - Fix the dependency conflict of geffnet installation ([#410](https://github.com/opendr-eu/opendr/pull/410))
+  - Fix bug in GEM ROS2 node ([#420](https://github.com/opendr-eu/opendr/pull/420))
+  - Fix link to nanodet documentation ([#421](https://github.com/opendr-eu/opendr/pull/421))
+  - EfficientLPS panoptic segmentation coloring bug ([#426](https://github.com/opendr-eu/opendr/pull/426))
+  - Bump flask from 1.1.2 to 2.3.2 ([#430](https://github.com/opendr-eu/opendr/pull/430))
+  - Fix tests on master branch ([#438](https://github.com/opendr-eu/opendr/pull/438))
+  - Added unzip installation as base ubuntu dependency and tool tests fixes ([#454](https://github.com/opendr-eu/opendr/pull/454))
+  - Active face recognition demo and bug fixes on Face Recognition ([#459](https://github.com/opendr-eu/opendr/pull/459))
+  - GPU installation fix ([#463](https://github.com/opendr-eu/opendr/pull/463))
+  - Fix ROS1 nodes argparse issue with .launch files ([#465](https://github.com/opendr-eu/opendr/pull/465))
+  - Minor fix on yolov5 webcam demo ([#466](https://github.com/opendr-eu/opendr/pull/466))
+  - Apply cuDNN init fix to all Object Detectors 2D ([#469](https://github.com/opendr-eu/opendr/pull/469))
+  - Updated test_suite_develop.yml based on latest test_suite.yml ([#471](https://github.com/opendr-eu/opendr/pull/471))
+  - Fix fmpgmapping ([#472](https://github.com/opendr-eu/opendr/pull/472))
+  - Synchronization and bugfixes ([#478](https://github.com/opendr-eu/opendr/pull/478))
+
 ## Version 2.2.0
 Released on July, 3rd, 2023.
 

@@ -1,6 +1,6 @@
 FROM ubuntu:20.04
 
-ARG branch=develop
+ARG branch=master
 ARG ros_distro=noetic
 
 # Install dependencies

@@ -1,6 +1,6 @@
 FROM nvidia/cuda:11.2.2-cudnn8-devel-ubuntu20.04
 
-ARG branch=develop
+ARG branch=master
 ARG ros_distro=noetic
 
 # Fix NVIDIA CUDA Linux repository key rotation

@@ -186,7 +186,7 @@ RUN echo 'source /opt/ros/${ROS_DISTRO}/setup.bash' >> /root/.bashrc
 CMD ["bash"]
 WORKDIR /
 
-RUN git clone --depth 1 --recurse-submodules -j8 --branch develop https://github.com/opendr-eu/opendr
+RUN git clone --depth 1 --recurse-submodules -j8 --branch master https://github.com/opendr-eu/opendr
 RUN apt-get update
 RUN cd ./opendr && ./bin/install_nvidia.sh $device
 RUN cd ./opendr/projects/opendr_ws/src && \

@@ -1,6 +1,6 @@
 # OpenDR Toolkit Reference Manual
 
-*Release 2.2.0*
+*Release 3.0.0*
 <div align="center">
   <img src="images/opendr_logo.png" />
 </div>

@@ -106,14 +106,14 @@ Note that `opendr-toolkit` is actually just a metapackage that includes all the
 ## CPU docker
 After installing [docker](https://docs.docker.com/engine/install/ubuntu/), you can directly run the OpenDR image as:
 ```bash
-sudo docker run -p 8888:8888 opendr/opendr-toolkit:cpu_v2.2.0
+sudo docker run -p 8888:8888 opendr/opendr-toolkit:cpu_v3.0.0
 ```
 The docker automatically runs a Jupyter notebook server that listens at port 8888.
 When launched, you can access the Jupyter notebook by following the link provided in the console, it should be similar to [http://127.0.0.1:8888/?token=TOKEN](http://127.0.0.1:8888/?token=TOKEN). In order to stop the container, please quit the Jupyter notebook.
 
 If you do not wish to use Jupyter, you can also experiment by starting an interactive session by running:
 ```bash
-sudo docker run -it opendr/opendr-toolkit:cpu_v2.2.0 /bin/bash
+sudo docker run -it opendr/opendr-toolkit:cpu_v3.0.0 /bin/bash
 ```
 In this case, do not forget to enable the virtual environment with:
 ```bash
@@ -122,18 +122,18 @@ source bin/activate.sh
 If you want to display GTK-based applications from the Docker container (e.g., visualize results using OpenCV `imshow()`), then you should mount the X server socket inside the container, e.g.,
 ```bash
 xhost +local:root
-sudo docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY opendr/opendr-toolkit:cpu_v2.2.0 /bin/bash
+sudo docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY opendr/opendr-toolkit:cpu_v3.0.0 /bin/bash
 ```
 
 ## GPU docker
 If you want to use a CUDA-enabled container please install [nvidia-docker](https://github.com/NVIDIA/nvidia-docker).
 Then, you can directly run the latest image with the command:
 ```bash
-sudo docker run --gpus all -p 8888:8888 opendr/opendr-toolkit:cuda_v2.2.0
+sudo docker run --gpus all -p 8888:8888 opendr/opendr-toolkit:cuda_v3.0.0
 ```
 or, for an interactive session:
 ```bash
-sudo docker run --gpus all -it opendr/opendr-toolkit:cuda_v2.2.0 /bin/bash
+sudo docker run --gpus all -it opendr/opendr-toolkit:cuda_v3.0.0 /bin/bash
 ```
 In this case, do not forget to enable the virtual environment with:
 ```bash
@@ -147,28 +147,61 @@ To install the toolkit, please first make sure that you have `git` available on
 ```bash
 sudo apt install git
 ```
-Then, clone the toolkit:
+
+### Clone the toolkit
+
+Then, shallow clone the toolkit:
 ```bash
 git clone --depth 1 --recurse-submodules -j8 https://github.com/opendr-eu/opendr
 ```
 
-If you want to install GPU-related dependencies, then you can appropriately set the `OPENDR_DEVICE` variable.
-The toolkit defaults to using CPU.
-Therefore, if you want to use GPU, please set this variable accordingly *before* running the installation script:
+or if you intend to use it as a development environment clone it fully:
+```bash
+git clone --recurse-submodules -j8 https://github.com/opendr-eu/opendr
+```
+
+### Installing for GPU
+
+If you want to install GPU-related dependencies, then you can appropriately set the `OPENDR_DEVICE` variable and install some 
+additional packages, otherwise skip this section. 
+
+To use GPU-enabled functionalities, you are strongly advised to install the following:
+
+1. NVIDIA drivers >460.106.00:
+2. CUDA 11.2: 
+   1. cuda_11.2.0_460.27.04_linux.run from [here](https://developer.nvidia.com/cuda-11.2.0-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=2004&target_type=runfilelocal)
+   2. Linux -> x86_64 -> Ubuntu-> 20.04 -> runfile local  
+   3. Skip driver installation, since drivers are already installed in the previous step
+3. cuDNN installed through [here](https://developer.nvidia.com/rdp/cudnn-archive)
+   1. Look for and download cuDNN v8.1.0 (January 26th, 2021), for CUDA 11.0, 11.1 and 11.2
+   2. cuDNN Runtime Library for Ubuntu20.04 x86_64 (Deb)
+
+The toolkit defaults to using CPU. Therefore, if you want to use GPU, please set this variable accordingly 
+*before* running the installation script:
+
 ```bash
 export OPENDR_DEVICE=gpu
 ```
 
-If you want to use ROS or ROS2, then you need to set the `ROS_DISTRO` variable *before* running the installation script so that additional required dependencies are correctly installed.
+**NOTE:** `OPENDR_DEVICE` does not alter the inference/training device at *runtime*.
+It only affects the dependency installation.
+You can use OpenDR API to change the inference device.
+
+### Using ROS
+
+If you want to use ROS or ROS2, then you need to set the `ROS_DISTRO` variable *before* running the installation script so that 
+additional required dependencies are correctly installed, otherwise skip this section.
 This variable should be set to either `noetic` or `melodic` for ROS, and `foxy` or `humble` for ROS2.
 
+### Installing the cloned repository
+
 You are then ready to install the toolkit:
 ```bash
 cd opendr
 ./bin/install.sh
 ```
 The installation script automatically installs all the required dependencies.
-Note that this might take a while (~10-20min depending on your machine and network connection), while the script also makes system-wide changes.
+Note that this might take a while (~10-20 minutes depending on your machine and network connection), while the script also makes system-wide changes.
 Using dockerfiles is strongly advised (please see below), unless you know what you are doing.
 Please also make sure that you have enough RAM available for the installation (about 4GB of free RAM is needed for the full installation/compilation).
 
@@ -180,23 +213,19 @@ source ./bin/activate.sh
 ```
 Then, you are ready to use the toolkit!
 
-**NOTE:** `OPENDR_DEVICE` does not alter the inference/training device at *runtime*.
-It only affects the dependency installation.
-You can use OpenDR API to change the inference device.
+### Verify the installation
 
 You can also verify the installation by using the supplied Python and C unit tests:
 ```bash
 make unittest
 make ctests
 ```
 
-If you plan to use GPU-enabled functionalities, then you are advised to install [CUDA 11.2](https://developer.nvidia.com/cuda-11.2.0-download-archive), along with [CuDNN](https://developer.nvidia.com/cudnn).
-
 **HINT:** All tests probe for the `TEST_DEVICE` enviromental variable when running.
 If this enviromental variable is set during testing, it allows for easily running all tests on a different device (e.g., setting `TEST_DEVICE=cuda:0` runs all tests on the first GPU of the system).
 
 
-## Nvidia embedded devices docker
+# Nvidia embedded devices docker
 You can also run the corresponding docker image on an Nvidia embedded device (supported: TX-2, Xavier-NX and AGX):
 
 Note that the embedded device should be flashed with Jetpack 4.6.

@@ -1,5 +1,6 @@
 engine
 perception/speech_recognition
+perception/speech_transcription
 perception/semantic_segmentation
 perception/binary_high_resolution
 perception/face_recognition
@@ -15,6 +16,7 @@ perception/object_detection_2d
 perception/object_tracking_2d
 perception/object_detection_3d
 perception/object_tracking_3d
+perception/gesture_recognition
 perception/panoptic_segmentation
 simulation/human_model_generation
 utils/hyperparameter_tuner

@@ -1,7 +1,7 @@
 <?xml version="1.0"?>
 <package format="2">
   <name>opendr_bridge</name>
-   <version>2.2.0</version>
+   <version>3.0.0</version>
   <description>OpenDR ROS bridge package. This package provides a way to translate ROS messages into OpenDR data types
       and vice versa.
   </description>

@@ -1,7 +1,7 @@
 <?xml version="1.0"?>
 <package format="2">
   <name>opendr_data_generation</name>
-  <version>2.2.0</version>
+  <version>3.0.0</version>
   <description>OpenDR's ROS nodes for data generation package</description>
   <maintainer email="tefas@csd.auth.gr">OpenDR Project Coordinator</maintainer>
   <license>Apache License v2.0 </license>