A Robotic Grasping State Perception Framework with Multi-Phase Tactile Information and Ensemble Learning
A Robotic Grasping State Perception Framework with Multi-Phase Tactile Information and Ensemble Learning
复制标题
具有多阶段触觉信息和集成学习的机器人抓取状态感知框架
DOI：
10.1109/lra.2022.3151260
复制
发表时间：
2022
影响因子：
5.2
通讯作者：
Shigeki Sugano
中科院分区：
文献类型：
--
作者：
Gang Yan;Alexander Schmitz;Satoshi Funabashi;Sophon Somlor;Tito Pradhono Tomo;Shigeki Sugano
近年来，触觉传感在机器人操作中越来越受到重视。在机器人触觉操作中，预测物体举起前的抓握稳定性和检测物体举起后正在发生或开始发生的滑移是两项重要且被广泛研究的任务。以前的方法侧重于针对上述两个任务之一提出新的神经网络（NN）架构，而没有考虑到这两个任务是在两个相互关联的动作阶段中使用的，即抓取和提升。因此，我们首先探索构建多相、多输出框架的可能性，将提升前的稳定性预测与提升后的滑移检测相结合。此外，为了提高预测/检测精度，我们还提出了使用各种方法（包括注意机制）显式集成不同的神经网络架构。我们的实验是在两个数据集上使用6种最先进的神经网络架构完成的，这些数据集包括3000多个机器人对80多个物体的抓取。此外，我们的建议在一个不可见物体的实时机器人实验中得到了验证。实验结果表明，所提出的多相、多输出模型比单相模型具有更高的可靠性和灵活性。我们还表明，使用不同神经网络架构的集成是提高整体性能的可行和实用的选择。
Recently, tactile sensing has attracted increasing attention for robotic manipulation. Predicting the grasping stability before lifting objects and detecting the ongoing/onset of slip after lifting objects are two critical and widely studied tasks in robotic tactile manipulation. Previous methods focus on proposing novel neural networks (NN) architectures towards one of the above two tasks and did not consider that the two tasks are employed in two interconnected action-phases, i.e. grasping and lifting. Therefore, we firstly explore the possibility of constructing a multi-phase, multi-output framework to combine the stability prediction before lifting and the slip detection after lifting. Moreover, to improve the prediction/detection accuracy, we also proposed to explicitly ensemble different NN architectures using various methods, including attention mechanisms. Our experiments are done with 6 state-of-art NN architectures on two datasets including more than 3000 robotic grasps over 80 objects in total. Furthermore our proposals are tested in a real-time robot experiment with unseen objects. Our experimental results show that the proposed multi-phase, multi-output model exhibits more reliable and flexible performance than a single phase model. We also show that using the ensemble of different NN architectures is a viable and practical choice to boost the overall performance.