Nov 20 - Munich AI, ML and Computer Vision Meetup

Join us for an evening of networking, talks, food and beverages at Macromedia University!

Date and Time

Nov 20, 2024 from 5:30 PM to 8:30 PM

Location

The Meetup will take place at Macromedia University, Sandstraße 9, Auditorium 608 in Munich

Robust and Efficient Coupling of Perception to Actuation with Metric and Non-Metric Scene Representations

We will present our work on robust and efficient ways, how camera information can be used to control an autonomous car. While many metric approaches require repeated intrinsic (sensor itself) and extrinsic (position of the sensor in the vehicle), our methods allow a direct control of the system solving the control problem directly in the retina (image plane) of the sensor. We will discuss the rich information encoded in the geometric projection of the scene that can be used for accurate navigation that includes uncertainty information necessary for fusion of diverse sensors and a novel way, how the dynamic information can be represented in the scene for planning in dynamic environments.

About the Speaker

Prof. Burschka of the Technical University of Munich conducts research into sensor systems in robotics and human-machine interfaces. Video-based navigation is one of his particular interests. This involves simulation of complex sensor systems through the analysis and fusion of sensor properties of physical sensor units and 3D reconstruction from the fusion of multimodal sensor data.

Understanding Context in the Wild - AI Testing Automated Driving Systems

Detailed contextual understanding is crucial for the testing of Automated Driving Systems (ADS). But, to provide high-quality and safe ADS, unseen events and potential weak points of the system need to be identified in the domain to be mitigated. This talk focuses on automated pipelines to identify the context, make driving data searchable and uncover potential weak points of driving systems automatically in the field.

About the Speaker

Tin Stribor Sohn is a doctoral candidate and tech lead for vehicle data analytics at Porsche AG, dealing with scenario search and failure cause analysis for automated driving. Tin holds an MSc. in computer science and co-founder/CTO of an energy startup.

Strategies Towards Reliable Scene Understanding for Autonomous Driving and Beyond

Scene understanding is crucial for self-driving cars and autonomous agents to reliably perceive their surroundings in diverse, unpredictable conditions. This talk tackles these challenges by presenting a series of research papers that improve reliability through novel methods in 2D and 3D perception, focusing on robustness in extreme scenarios and generalization to unseen environments.

About the Speaker

Stefano Gasperini is the Co-Founder & CEO of Visualais, a Computer Vision startup enabling the creation of 3D renderings from smartphone images. Beyond numerous top-tier research papers and outstanding reviewer awards from his PhD at TUM, as PostDoc, Stefano keeps contributing to Computer Vision and AI by advising a team of PhD students at TUM.

How to Unlock More Value from Self-Driving Datasets

AV/ADAS is one of the most advanced fields in Visual AI. However, getting your hands on a high quality dataset can be tough, let alone working with them to get a model to production. In this talk, I will show you the leading methods and tools to help visualize as well take these datasets to the next level. I will demonstrate how to clean and curate AV datasets as well as perform state of the art augmentations using diffusion models to create synthetic data that can empower the self driving car models of the future,

About the Speaker

Daniel Gural is a seasoned Machine Learning Engineer at Voxel51 with a strong passion for empowering Data Scientists and ML Engineers to unlock the full potential of their data.