EchoVPR: Echo State Networks for Visual Place Recognition
EchoVPR: Echo State Networks for Visual Place Recognition
复制标题
DOI：
10.1109/lra.2022.3150505
复制
发表时间：
2021-10
影响因子：
5.2
通讯作者：
Anil Özdemir;A. Barron;Andrew O. Philippides;M. Mangan;E. Vasilaki;Luca Manneschi
中科院分区：
文献类型：
--
作者：
Anil Özdemir;A. Barron;Andrew O. Philippides;M. Mangan;E. Vasilaki;Luca Manneschi
识别以前访问过的位置是自主导航中一个重要但尚未解决的任务。当前的视觉位置识别（VPR）基准测试通常要求模型从包含空间和时间组件的序列数据集中恢复查询图像（或图像）的位置。最近，回声状态网络（ESN）被证明在解决需要时空建模的机器学习任务方面特别强大。这些网络是简单但功能强大的神经结构，它们在多个时间尺度和非线性高维表示上表现出记忆，可以发现数据中的时间关系，同时仍然保持学习中的线性。本文给出了一系列的ESNs，并分析了它们在VPR问题中的适用性。我们报告说，在六个标准基准测试（GardensPoint、speedtest、ESSEX3IN1、Oxford RobotCar和Nordland）中，将ESNs添加到预处理卷积神经网络中，与非循环网络相比，性能有了显著提高，这表明ESNs能够捕获VPR问题中固有的时间结构。此外，我们表明，包含esn的模型可以优于同类领先的VPR模型，后者也利用了数据的顺序动态。最后，我们的研究结果表明，esn还提高了泛化能力、鲁棒性和准确性，进一步支持了它们对VPR应用的适用性。
Recognising previously visited locations is an important, but unsolved, task in autonomous navigation. Current visual place recognition (VPR) benchmarks typically challenge models to recover the position of a query image (or images) from sequential datasets that include both spatial and temporal components. Recently, Echo State Network (ESN) varieties have proven particularly powerful at solving machine learning tasks that require spatio-temporal modelling. These networks are simple, yet powerful neural architectures that - exhibiting memory over multiple time-scales and non-linear high-dimensional representations - can discover temporal relations in the data while still maintaining linearity in the learning. In this paper, we present a series of ESNs and analyse their applicability to the VPR problem. We report that the addition of ESNs to pre-processed convolutional neural networks led to a dramatic boost in performance in comparison to non-recurrent networks in five out of six standard benchmarks (GardensPoint, SPEDTest, ESSEX3IN1, Oxford RobotCar, and Nordland) demonstrating that ESNs are able to capture the temporal structure inherent in VPR problems. Moreover, we show that models that include ESNs can outperform class-leading VPR models which also exploit the sequential dynamics of the data. Finally, our results demonstrate that ESNs also improve generalisation abilities, robustness, and accuracy further supporting their suitability to VPR applications.