双曲グラフ畳み込みネットワークを用いた半教師ありベクトル量子化によるVisual SLAM
Semi-supervised Vector-Quantization in Visual SLAM using HGCN
Visual SLAMのループ閉じ込め検出において、双曲グラフ畳み込みネットワーク(HGCN)でSURF特徴のベクトル量子化を半教師ありで行う手法を提案し、ORB-SLAMへの拡張も示した。
著者: Amir Zarringhalam, Saeed Shiry Ghidary, Ali Mohades Khorasani
分類: cs.CV, cs.RO
原文アブストラクト
In this paper, two semi-supervised appearance based loop closure detection technique, HGCN-FABMAP and HGCN-BoW are introduced. Furthermore an extension to the current state of the art localization SLAM algorithm, ORB-SLAM, is presented. The proposed HGCN-FABMAP method is implemented in an off-line manner incorporating Bayesian probabilistic schema for loop detection decision making. Specifically, we let a Hyperbolic Graph Convolutional Neural Network (HGCN) to operate over the SURF features graph space, and perform vector quantization part of the SLAM procedure. This part previously was performed in an unsupervised manner using algorithms like HKmeans, kmeans++,..etc. The main Advantage of using HGCN, is that it scales linearly in number of graph edges. Experimental results shows that HGCN-FABMAP algorithm needs far more cluster centroids than HGCN-ORB, otherwise it fails to detect loop closures. Therefore we consider HGCN-ORB to be more efficient in terms of memory consumption, also we conclude the superiority of HGCN-BoW and HGCN-FABMAP with respect to other algorithms.
関連論文
- ブラケティングで道を切り開く:ロバストなVisual SLAMのための適応的マルチ露出ポリシーVisual SLAM
- 線特徴拡張cuVSLAM:低テクスチャ環境における堅牢な視覚オドメトリVisual SLAM
- ヒューマン・イン・ザ・ループ型ニューロシンボリックドリフト予測による信頼性の高いVisual SLAMVisual SLAM
- Know-Your-Scene (KYS)-SLAM: ステレオ視覚SLAMにおける特徴マッチングのための階層的意味・運動事前分布Visual SLAM
- HuMemSLAM: 人間の記憶に着想を得た効率的な意味的場所認識による堅牢なVisual SLAMVisual SLAM
- 深層単眼Visual SLAMはスケール問題を克服したのか?ScaleMasterデータセットとベンチマークVisual SLAM
