ViSaRL: Saliency-Guided Visual Representations for Robot Learning

  • Author By
  • Publication date October 7, 2024
  • Categories: Blog
  • Sticky post

By Anthony Liang Teaching robots to perform complex control tasks from high-dimensional image inputs is a nontrivial problem. State-of-the-art methods require hundreds of manually collected expert demonstrations or hours to days of training in simulation to learn a simple pick-and-place skill. One reason for this is that images are cluttered with distractions for robots, such…Continue Reading ViSaRL: Saliency-Guided Visual Representations for Robot Learning

From Generalists to Specialists: A Case for Real-World RL in Robot Manipulation

  • Author By
  • Publication date September 24, 2025
  • Categories: Blog

A USC RASC blog post by Jesse Zhang and Abrar Anwar. Introduction In The Automatic Motorist, a robot chauffeur takes a newlywed couple from Earth’s dirt roads to Saturn’s rocky rings—all while evading police. This 1911 (!) short film depicted a generalist robot before the word “robot” even existed. Until recently, robotics research focused on…Continue Reading From Generalists to Specialists: A Case for Real-World RL in Robot Manipulation

Capturing the Data That Users Naturally Emit

The widespread success of foundation models like ChatGPT, Gemini, Deepseek, and Claude demonstrates the effectiveness of scaling data collection to address many vision and language tasks. Robotics has also seen benefits from learning from large amounts of high-quality data, but robots are still not reliably performing many of the tasks that people do every day…Continue Reading Capturing the Data That Users Naturally Emit

USC at ICRA 2025

USC School of Advanced Computing and USC Viterbi researchers showcase breakthroughs in generative modeling, safety in imitation learning, human-aware planning and more at ICRA 2025, one of the most prestigious gatherings in robotics….Continue Reading USC at ICRA 2025

MILE: Model-based Intervention Learning

Imagine deploying a robot in a new environment. While it has been pre-trained in a factory, its performance is far from perfect when faced with real-world variability. Whether it’s a warehouse robot handling unexpected objects or a household assistant adapting to a cluttered kitchen, fine-tuning is often necessary to bridge the gap between training and…Continue Reading MILE: Model-based Intervention Learning