🔔 We are proud to announce a new publication supported by CIRCULess, advancing research in robotics, intelligent automation, and multi-modal perception!
As robotic systems increasingly operate in complex and changing environments, reliable perception is essential for safe and effective navigation. Based on this, our partners from Luleå University of Technology (LTU) have published the research paper “GenLaM: Generative Layered Mesh for Multi-Modal Sensor Emulation in Robotics.”
👉 The study introduces GenLaM, a framework that uses a single RGB image to generate detailed 3D scene representations and emulate multiple sensor outputs. The approach reduces dependence on multiple physical sensors while supporting robust robotic perception and navigation.
🔎 Highlights:
• Generates 3D scene representations from a single RGB image.
• Enables RGB and segmented LiDAR sensor emulation.
• Reconstructs challenging transparent surfaces, such as glass.
• Supports robotic path planning and obstacle avoidance.
• Evaluated under challenging environmental conditions, including low light, rain, fog, and snow.
👏 Thanks to all the contributors: Aakash Singh Bais, Akash Patel, Christoforos Kanellakis, and George Nikolakopoulos.
This research contributes to the development of more capable and adaptable robotic systems for complex real-world environments.
📖 Read the full publication here: https://events.infovaya.com/uploads/documents/pdfviewer/90/2d/220034-4358.pdf
#CIRCULess #HorizonEurope #Robotics #ArtificialIntelligence #Automation #Digitalisation #Research #Innovation
