Dark Reciprocal-Rank: Teacher-to-student Knowledge Transfer from Self-localization Model to Graph-convolutional Neural Network
Dark Reciprocal-Rank: Teacher-to-student Knowledge Transfer from Self-localization Model to Graph-convolutional Neural Network
复制标题
暗倒数秩：从自定位模型到图卷积神经网络的教师到学生的知识迁移
DOI：
10.1109/icra48506.2021.9561158
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Kanji Tanaka
中科院分区：
文献类型：
--
作者：
Koji Takeda;Kanji Tanaka
在视觉机器人的自定位中，基于图的场景表示和匹配作为鲁棒性强、判别性强的自定位方法近年来受到了广泛的关注。虽然有效，但它们的计算和存储成本不能很好地扩展到大型环境。为了缓解这一问题，我们将自定位表述为一个图分类问题，并尝试使用图卷积神经网络（GCN）作为图分类引擎。一种直接的方法是使用最先进的自定位系统所使用的视觉特征描述符，直接作为图节点特征。然而，它们在原始自定位系统中的优越性能不一定能在基于gcn的自定位系统中得到复制。为了解决这一问题，我们引入了一种新的基于等级匹配的师生知识转移方案，该方案使用现有的最先进的教师自我定位模型输出的倒数秩向量作为暗知识进行转移。实验表明，所提出的图卷积自定位网络（GCLN）可以显著优于最先进的自定位系统以及教师分类器。代码和数据集可从https://github.com/KojiTakeda00/Reciprocal_rank_KT_GCN获得。
In visual robot self-localization, graph-based scene representation and matching have recently attracted research interest as robust and discriminative methods for self-localization. Although effective, their computational and storage costs do not scale well to large-size environments. To alleviate this problem, we formulate self-localization as a graph classification problem and attempt to use the graph convolutional neural network (GCN) as a graph classification engine. A straightforward approach is to use visual feature descriptors that are employed by state-of-the-art self-localization systems, directly as graph node features. However, their superior performance in the original self-localization system may not necessarily be replicated in GCN-based self-localization. To address this issue, we introduce a novel teacher-to-student knowledge-transfer scheme based on rank matching, in which the reciprocal-rank vector output by an off-the-shelf state-of-the-art teacher self-localization model is used as the dark knowledge to transfer. Experiments indicate that the proposed graph-convolutional self-localization network (GCLN) can significantly outperform state-of-the-art self-localization systems, as well as the teacher classifier. The code and dataset are available at https://github.com/KojiTakeda00/Reciprocal_rank_KT_GCN.