Fleet Active Learning: A Submodular Maximization Approach
Fleet Active Learning: A Submodular Maximization Approach
复制标题
DOI：
--
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Oguzhan Akcin;Orhan Unuvar;Onat Ure;Sandeep P. Chinchali
中科院分区：
文献类型：
--
作者：
Oguzhan Akcin;Orhan Unuvar;Onat Ure;Sandeep P. Chinchali
在多机器人系统中，机器人经常收集数据以提高其深度神经网络（dnn）的感知和规划性能。理想情况下，这些机器人应该通过采用主动学习方法从其本地数据分布中选择最具信息量的样本。然而，当数据收集分布在多个机器人之间时，由于不同的机器人可能会选择相似的数据点，因此冗余成为一个问题。为了克服这一挑战，我们提出了一个车队主动学习（FAL）框架，在该框架中，机器人集体选择信息丰富的数据样本来增强其DNN模型。我们的框架利用子模块最大化技术来优先选择具有高信息增益的样本。通过迭代算法，机器人协调它们的努力，共同选择最有价值的样本，同时最大限度地减少机器人之间的通信。我们对我们提出的框架的性能进行了理论分析，并表明它能够近似NP-hard最优解。我们通过在真实世界的感知和分类数据集（包括Berkeley DeepDrive等自动驾驶数据集）上的实验证明了我们的框架的有效性。我们的结果显示，提高了25%。分类准确率为0%，9。平均精度为2%。与完全分布的基线相比，子模块目标值降低了5%。
: In multi-robot systems, robots often gather data to improve the performance of their deep neural networks (DNNs) for perception and planning. Ideally, these robots should select the most informative samples from their local data distributions by employing active learning approaches. However, when the data collection is distributed among multiple robots, redundancy becomes an issue as different robots may select similar data points. To overcome this challenge, we propose a fleet active learning (FAL) framework in which robots collectively select informative data samples to enhance their DNN models. Our framework leverages submodular maximization techniques to prioritize the selection of samples with high information gain. Through an iterative algorithm, the robots coordinate their efforts to collectively select the most valuable samples while minimizing communication between robots. We provide a theoretical analysis of the performance of our proposed framework and show that it is able to approximate the NP-hard optimal solution. We demonstrate the effectiveness of our framework through experiments on real-world perception and classification datasets, which include autonomous driving datasets such as Berkeley DeepDrive. Our results show an improvement by up to 25 . 0% in classification accuracy, 9 . 2% in mean average precision and 48 . 5% in the submodular objective value compared to a completely distributed baseline.