In the past few years, open source frameworks such as PyTorch have greatly accelerated the development of research in AI subareas such as computer vision and natural language processing. These frameworks provide researchers with a level of abstraction that makes it easier to build state-of-the-art systems, use shared libraries and tools, and optimize performance. So to provide similar utility in ai robotics research, Facebook teamed up with researchers at Carnegie Mellon University to build and open up PyRobot, a robotics framework.
At the same time, it reduces the barriers to entry for AI researchers. Without having to specialize in the details of hardware, drivers and control systems, researchers can easily set up and run the robot in just a few hours. If PyRobot is used together with LoCoBot and other low-cost robot platforms, PyRobot can greatly reduce the cost of robot research, which is more conducive to the promotion and promotion of robot research.
Facebook, on the other hand, wants To make PyRobot an ecosystem for robotics research. PyRobot not only makes it easier for community Zhongbo Love Platform researchers to use robot datasets, algorithm implementations and models, but also helps them set benchmarks and share data.
PyRobot currently supports robots Sawyer and LoCoBot, while more support models are coming and are being integrated with simulators such as MuJoCo and Habitat.