Robotics

MeMo: a novel framework for pretraining meaningful, modular robot controllers via noise injection, enabling efficient transfer learning across different robot morphologies and tasks.

Make-An-Agent generates high-performing robotic control policies from single behavioral demonstrations using behavior-prompted diffusion, showcasing impressive generalization and real-world applicabil…

LLaMAR: LM-based planner for multi-agent robots excels in long-horizon, partially observable tasks, achieving 30% higher success than existing methods.

LLM-based Skill Diffusion (LDuS) enables zero-shot robotic policy adaptation to various contexts specified in natural language by generating controllable skill trajectories via loss-guided diffusion a…

SDF-Sim: A novel learned rigid-body simulator that leverages SDFs to achieve unprecedented scalability, enabling simulations with hundreds of objects and millions of nodes.

Actionable AI agents are trained efficiently via a novel framework, VPDD, which uses discrete diffusion to pre-train on massive human videos, and fine-tunes on limited robot data for superior multi-ta…

iVideoGPT: A scalable, interactive world model trained on millions of human & robot manipulation videos, enabling efficient video prediction and model-based reinforcement learning.

Robots using LLMs for task planning often make unsafe or wrong decisions due to LLM hallucination and ambiguity in instructions. This paper introduces ‘introspective planning,’ a novel method that us…

Stable closed-loop control in latent space is achieved using a novel Coupled Oscillator Network, offering efficient model-based control for complex nonlinear systems directly from image data.

Humanoid robots now walk in San Francisco zero-shot, thanks to a novel ’next token prediction’ approach trained on diverse sensorimotor data, enabling real-world generalization and data efficiency.

Robots can now dexterously grasp objects based on natural language commands thanks to DexGYS, a new language-guided dexterous grasp generation framework and dataset.

GOOSE: a novel framework using graph learning for efficient and interpretable numeric planning, outperforming existing methods in many benchmarks.

GarmentLab: A new benchmark and simulation platform tackles garment manipulation challenges by offering realistic simulations, diverse assets, and tasks bridging the sim-to-real gap for more robust AI…

FACTORSim generates full, coded simulations from natural language descriptions, outperforming existing methods in accuracy and zero-shot transfer learning by using a factored POMDP representation.

EGODE, a novel framework, leverages coupled graph ODEs and an event module to accurately model continuous and instantaneous changes in rigid body dynamics, outperforming existing methods.

DynaMo: a novel self-supervised method significantly boosts visuo-motor control by learning in-domain dynamics from limited expert demonstrations, improving policy performance across various environme…

DISC, a novel control policy, disentangles untrusted ML predictions to achieve near-optimal performance when accurate, while guaranteeing competitive ratio bounds even with significant prediction erro…

DP-Attacker unveils diffusion-based policy vulnerabilities by crafting effective adversarial attacks, significantly impacting robot safety and paving the way for more robust AI.

Diffusion Forcing merges next-token prediction and full-sequence diffusion for superior sequence generation.

DiffuserLite: a super-fast diffusion planning framework achieving real-time performance (122Hz).