BeamNG presented Generating Artificial Datasets at Deep Layers Workshop 2022.
BeamNG presented its new semantic annotation dataset for driving. It contains 50000 frames captured during more than 25 hours of driving throughout two different maps in BeamNG. The dataset was generated using the capabilities of BeamNGpy Python library and the camera sensor, which is a part of the BeamNG.tech suite. The dataset also contains metadata for every frame containing the position and state of the ego-vehicle.
The dataset will be released later in the month.
About Deep Layers:
Deep Layers forum serves as a place to connect people, share knowledge, compare approaches and see how specific model architectures can help to solve particular tasks. The meeting was organized by a scientific group of Artificial Intelligence and Medical Technologies with the support of the Brno.AI cluster and with the finnacial support of Strategie AV21.
The program for Deep Layers 2022 can be downloaded here.
September 20th - Machine learning tutorials (in Python language)
- Morning blocks: Introduction to the machine learning, how to treat & check your data, fast & easy-to-use ML methods
- Afternoon blocks: Introduction to Deep Learning for newbies: layers types, basic architectures and final DL tutorial
September 21st - Machine learning applications
- Morning block 1: Deep learning in Actual Life
- Poster session
- Morning block 2: Deep learning & neuroscience
- Afternoon blocks: Deep learning in image classification and image reconstruction