imitation learning
PulseAugur coverage of imitation learning — every cluster mentioning imitation learning across labs, papers, and developer communities, ranked by signal.
8 day(s) with sentiment data
-
New ACT-JEPA architecture enhances AI policy representation learning
Researchers have developed ACT-JEPA, a novel architecture that combines imitation learning (IL) and self-supervised learning (SSL) to improve policy representation learning. This approach trains end-to-end to predict bo…
-
New RL framework uses language for adaptive guidance; survey covers LLM distillation techniques · 2 sources tracked
Researchers have introduced Hierarchical Reinforcement Learning with Language Instructions (HRLLI), a novel framework that enhances reinforcement learning efficiency by dynamically selecting relevant natural language gu…
-
New R2BC method enables single-human training of multi-robot systems
Researchers have developed Round-Robin Behavior Cloning (R2BC), a novel method for training multi-robot systems using sequential, single-agent demonstrations. This approach allows a single human operator to teach comple…
-
MIT Students Earn Hertz Fellowships for AI Research
Four individuals affiliated with MIT have been awarded prestigious Hertz Fellowships for their significant contributions to AI research. Their work spans areas including predictive models and imitation learning, highlig…
-
Robotics method uses RL to handle missing sensor data
Researchers have developed RL4IL, a novel reinforcement learning approach designed to enhance multimodal imitation learning in robotics, particularly when sensor data is missing. This method utilizes reinforcement learn…
-
New AI Policy Enhances Sidewalk Navigation with Human Preference Learning
Researchers have developed FlowPilot, a novel navigation policy for autonomous sidewalk navigation using only a monocular RGB camera. This system addresses limitations of traditional imitation learning by employing anch…
-
AI Research Focuses on Imperfect Feedback for Model Learning
AI researchers are now prioritizing the study of imperfect feedback in predictive models. This focus is particularly relevant for areas such as imitation learning and bandit problems, where models must learn from incomp…
-
AI imitation learning metrics fail to detect critical errors
Researchers have developed a controlled testbed to evaluate imitation-learning curation metrics, finding that action-only scoring methods fail to detect critical structural errors in demonstration data. These errors, wh…
-
Behavior Cloning approximates MPC for robotic manipulators
Researchers have explored using Behavior Cloning to create computationally efficient approximations of Model Predictive Control (MPC) policies for robotic manipulators. The study focused on a 3-degree-of-freedom manipul…
-
HOIST method enhances humanoid robot load manipulation
Researchers have developed a new method called HOIST to improve the ability of humanoid robots to manipulate suspended loads. This approach combines imitation learning from human demonstrations with sample-efficient rei…
-
Research paper explains PIDM's advantage over behavior cloning
A new research paper explores the effectiveness of Predictive Inverse Dynamics Models (PIDM) compared to Behavior Cloning (BC) for imitation learning. The study theoretically explains that PIDM offers a bias-variance tr…
-
New algorithm bridges imitation gap in reinforcement learning
Researchers have developed a new algorithm to address the imitation gap in reinforcement learning, particularly in robotics. The method focuses on creating a shared embedding space that prevents the teacher policy from …
-
COBALT platform uses smartphones for crowdsourced robot learning data
Researchers have developed COBALT, a cloud-based teleoperation platform that uses smartphones to collect large-scale robot learning data. This system allows multiple users to contribute to robot training simultaneously,…
-
Georgia Tech's SAIL system enables robots to learn and perform tasks 4x faster than humans
Researchers at Georgia Tech have developed a new system called SAIL (Speed Adaptation for Imitation Learning) that enables robots to perform tasks significantly faster than human instructors. Traditional imitation learn…
-
TAVIS benchmark advances robotics imitation learning with active vision
Researchers have introduced TAVIS, a new benchmark designed to evaluate active vision in imitation learning for robotics. The benchmark includes two task suites, TAVIS-Head and TAVIS-Hands, built on humanoid embodiments…
-
New LC-MAPF model enhances multi-agent pathfinding with local communication
Researchers have developed a new machine learning model called LC-MAPF designed to improve coordination in large-scale multi-agent pathfinding scenarios. This model incorporates a learnable communication module that all…
-
Researchers develop provable imitation learning for plasma control in fusion
Researchers have developed a theoretical framework for imitation learning to control instabilities in partially-observed Vlasov--Poisson equations, a key challenge in nuclear fusion. The method distills expert policies,…
-
AI framework SeqLight automates stage lighting control using imitation learning
Researchers have developed SeqLight, a deep learning framework designed to automate stage lighting control inspired by music. This hierarchical model maps music to the Hue-Saturation-Value (HSV) color space for multiple…
-
New framework evaluates autonomous driving AI robustness against real-world adversarial attacks
Researchers have developed a new framework for evaluating the real-time robustness of autonomous driving systems against adversarial attacks. This approach utilizes real-world intersection driving data, moving beyond pu…