LeRobot v0.6.0: Imagine, Evaluate, Improve
More from the blog
Looking beyond natural sequences
A new machine-learning framework aims to improve the success rate of computational protein design while moving away from results that reproduce sequences found in nature.
Record, train, and deploy from one place with Strands Agents, LeRobot, and Hugging Face Storage Buckets
A Blog post by Amazon on Hugging Face
Toward a future that preserves benefits of neurotechnology for all
PhD student Rachel Sava, winner of the Envisioning the Future of Computing Prize, explores transformative improvements and dystopian risks of neural technology.
From the Hugging Face Hub to robot hardware with Strands Agents and LeRobot
Related papers
Transformer-Based Autonomous Driving Models and Deployment-Oriented Compression: A Survey
This survey reviews representative Transformer-based autonomous driving models and organizes them by task role, sensing configuration, and architectural design and analyzes how efficiency constraints reshaping model design choices in practice affects deployability, robustness, and safety.
FlowCorrect: Efficient Interactive Correction of Generative Flow Policies for Robotic Manipulation
The results clearly demonstrate that FlowCorrect learns from very few demonstrations and enables fast, sample-efficient, incremental, human-in-the-loop corrections of generative visuomotor policies at deployment time in real-world robotics.
Spatial-Semantic Reasoning using Large Language Models for Efficient UAV Search Operations
A real-time semantic navigation framework for Unmanned Aerial Vehicles (UAVs) focused on improving time efficiency in the Object Goal Navigation (ObjectNav) task, using a Large Language Model that interprets user-provided natural language instructions and performs semantic reasoning over detected objects and spatial context to prioritize high-probability search regions.
SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Models
Experiments on simulated and real-world benchmarks demonstrate that SCALE improves state-of-the-art VLAs and outperforms existing TTS methods while maintaining single-pass efficiency.