FEATURED ROBOTICS
AI Learns to See in 3D Without Depth Sensor
From arXiv • Latest Research
We developed a new computer vision system called DropD-SLAM that can create accurate 3D maps using just one camera. This system uses pre-trained models to detect objects, track movement, and create detailed 3D maps, making it a promising step towards more affordable and efficient navigation systems.
Read Full Paper →
Rise Of The Intelligent Robots That Understand What We Want
Scientists created a new system that allows robots to understand and follow instructions without needing large amounts of labeled data or expensive training. This breakthrough has the potential to make robots more flexible, adaptable, and able to accomplish tasks in many different environments, which could lead to significant advancements in robotics and automation.
Read Paper →
AI Learns Hair Styles From Data
Scientists developed a robot system that can style hair without human help, using a combination of computer modeling and machine learning to understand how hair behaves. This innovation has the potential to make it easier for people with limited mobility to take care of their hair, and could also lead to more advanced robot hair-styling capabilities in the future.
Read Paper →
AI Gets Smarter on Robot Navigation with New Data-Driven System
We created a new motion planning system that uses powerful computers to quickly find safe paths for robots in complex environments. This system could help make robots more efficient and able to respond faster to changing situations, potentially leading to better performance and safety in various industries and applications.
Read Paper →
AI Learns from User Data to Understand Preferences Better
Scientists created a new version of a computer program that helps design and test underwater robots. This improved program has the potential to make it easier and more accurate to develop and use underwater robots, which could lead to breakthroughs in ocean exploration and other areas.
Read Paper →