AlejandroMllo/action_flow_matching
Code for the paper "Action Flow Matching for Continual Robot Learning" presented at Robotics: Science and Systems (RSS) 2025.
Action Flow Matching (AFM) helps robotics engineers and researchers refine robot models continuously as new data becomes available. It takes data on intended versus actual robot movements and outputs improved action plans, allowing robots to adapt more effectively to changing environments or tasks. This tool is designed for those working on autonomous systems that need to learn and improve their physical interactions in real-time.
Use this if you need to continuously improve a robot's ability to execute planned actions and adapt to discrepancies between its intended movements and actual outcomes, without needing to retrain the entire system from scratch.
Not ideal if you are looking for a general-purpose robot control framework or a tool for initial robot programming, as its focus is specifically on online model refinement for existing robotic systems.
Stars
31
Forks
1
Language
Python
License
MIT
Category
Last pushed
Dec 18, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/generative-ai/AlejandroMllo/action_flow_matching"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related tools
AngeClementAkazan/Sequential-FeatureForestFlow
This repository contains the code source of the Heterogeneous Sequential Feature Forest Flow...
dsgiitr/flux-watermarking
Official Implementation of the paper WMARK@ICLR Detection Limits and Statistical Separability of...
ericbill21/FOCUS
Official codebase for FOCUS: Optimal Control Meets Flow Matching: A Principled Route to...
lebellig/flow-matching
Annotated Flow Matching paper
lebellig/discrete-fm
Educational implementation of the Discrete Flow Matching paper