MTF-Grasp: A Multi-tier Federated Learning Approach for Robotic Grasping

2025-07-14

AI Summary

To address the challenges of non-IID data distribution and scarce samples in robotic grasping tasks—leading to degraded global model performance and limited privacy preservation—this paper proposes a multi-level collaborative federated learning framework. The framework innovatively categorizes robots into high-tier and low-tier roles: high-tier robots, equipped with abundant, high-quality data, generate a robust seed model; a data-distribution-aware targeted model dissemination strategy then guides local training on low-tier robots. Low-tier robots upload only model updates—ensuring raw data remain locally confined. Experiments on the Cornell and Jacquard datasets under label-skewed (quantity-skewed) non-IID settings demonstrate that our method achieves up to an 8.0% improvement in grasping accuracy over baselines such as FedAvg. It effectively mitigates non-IID-induced model degradation while simultaneously ensuring strong privacy guarantees and efficient cross-device collaboration.

Technology Category

• Intelligent Robots: Manipulation
• Machine Learning: Distributed Machine Learning & Federated Learning
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated Learning (FL) is a promising machine learning paradigm that enables participating devices to train privacy-preserved and collaborative models. FL has proven its benefits for robotic manipulation tasks. However, grasping tasks lack exploration in such settings where robots train a global model without moving data and ensuring data privacy. The main challenge is that each robot learns from data that is nonindependent and identically distributed (non-IID) and of low quantity. This exhibits performance degradation, particularly in robotic grasping. Thus, in this work, we propose MTF-Grasp, a multi-tier FL approach for robotic grasping, acknowledging the unique challenges posed by the non-IID data distribution across robots, including quantitative skewness. MTF-Grasp harnesses data quality and quantity across robots to select a set of "top-level" robots with better data distribution and higher sample count. It then utilizes top-level robots to train initial seed models and distribute them to the remaining "low-level" robots, reducing the risk of model performance degradation in low-level robots. Our approach outperforms the conventional FL setup by up to 8% on the quantity-skewed Cornell and Jacquard grasping datasets.

Problem

• Addresses non-IID data in robotic grasping with FL
• Improves model performance in low-data robots
• Ensures privacy while training collaborative grasping models

Innovation

• Multi-tier FL for robotic grasping
• Top-level robots train seed models
• Improves performance on non-IID data