D5.3 – Report on assessment of Hybrid AI
Deliverable D5.3 elaborates on the end to end Trustworthiness including ethics. Methodology, implementation, and evaluation results . To be published
D6.5 Mid-term workshop report
Deliverable 6.5 summarizes the main ULTIMATE technical contributions till the half of the project whilegathering the UC owners and ULTIMATE Advisory Board feedback (from the M12 technicalreview and a dedicated workshop on 18/04/2024) to provide the reader with an end-user pointof view. It eases the acceptance of the ULTIMATE hybrid AI solutions. Download the deliverable
ULTIMATE project Youtube Channel
Discover our demos and technical videos on our Youtube Channel ! ULTIMATE Youtube Channel
ULTIMATE project final poster
The ULTIMATE poster with the results of the project is available. Inside you can find the works, use-case to demo and the results for the project. Download the poster here
ULTIMATE WORKSHOP UC SYNTHETIC DATASET V1 (ROS 2 – HUMBLE)
The ULTIMATE WORKSHOP UC SYNTHETIC DATASET V1 dataset is composed of 24 files (12 ROS1-Noetic and 12 ROS2-Humble recordings), each one containing a scene reproducing a hybrid environment populated by 3 RGBD cameras, two humans and robots. The included human animations reproduce the ITTI Workshop UC, which includes 7 particular poses, 2 normal activities and 4 dangerous […]
ULTIMATE WORKSHOP UC SYNTHETIC DATASET V1 (ROS1 NOETIC)
The ULTIMATE WORKSHOP UC SYNTHETIC DATASET V1 dataset is composed of 24 files (12 ROS1-Noetic and 12 ROS2-Humble recordings), each one containing a scene reproducing a hybrid environment populated by 3 RGBD cameras, two humans and robots. The included human animations reproduce the ITTI Workshop UC, which includes 7 particular poses, 2 normal activities and 4 dangerous […]
Rohucad Data Set
RoHuCAD is a dataset of human-robot collaboration in a robotic workshop (check workshop_layout.png). Two robots (collaborative manipulator – cobot, autonomous mobile robot – AMR) assist three human operators in assembly of electronic devices. There are two 8-min long recordings in the dataset. They mostly follow the same scenario, with slightly different anomalies. The data is in […]
tf2-bfgs
tf2-bfgs is a wrapper to train a tf.Module or tf.keras.Model using (L-)BFGS optimizer from Tensorflow Probability. MIT License Copyright (c) 2024 Matthieu Barreau github repository
scio
scio: confidence scores for Neural Networks, made easy! If you are an ML engineer, research scientist or amateur programmer, this Python library is meant for you! Use scio to augment your PyTorch Neural Network predictions with confidence scores. It applies seamlessly to trained models, facilitating collaborative work and real-world deployment! Github repository: github.com/ThalesGroup/scio Documentation: scio.readthedocs.io […]
ULTIMATE INDUSTRY UC SYNTHETIC DATASET V1
The ULTIMATE INDUSTRIAL UC SYNTHETIC DATASET V1 is composed of 15 folders, each one containing a picture frame reproducing a hybrid environment populated by1 RGBD camera. The 3D scene is a digital twin of the real scene of the Industry UC. The captures are extensively annotated in each folder. Creative Commons Attribution 4.0 International Zenodo […]