{"title":"ORB-SfMLearner:具有选择性在线适应功能的 ORB 引导的自监督视觉测距仪","authors":"Yanlin Jin, Rui-Yang Ju, Haojun Liu, Yuzhong Zhong","doi":"arxiv-2409.11692","DOIUrl":null,"url":null,"abstract":"Deep visual odometry, despite extensive research, still faces limitations in\naccuracy and generalizability that prevent its broader application. To address\nthese challenges, we propose an Oriented FAST and Rotated BRIEF (ORB)-guided\nvisual odometry with selective online adaptation named ORB-SfMLearner. We\npresent a novel use of ORB features for learning-based ego-motion estimation,\nleading to more robust and accurate results. We also introduce the\ncross-attention mechanism to enhance the explainability of PoseNet and have\nrevealed that driving direction of the vehicle can be explained through\nattention weights, marking a novel exploration in this area. To improve\ngeneralizability, our selective online adaptation allows the network to rapidly\nand selectively adjust to the optimal parameters across different domains.\nExperimental results on KITTI and vKITTI datasets show that our method\noutperforms previous state-of-the-art deep visual odometry methods in terms of\nego-motion accuracy and generalizability.","PeriodicalId":501130,"journal":{"name":"arXiv - CS - Computer Vision and Pattern Recognition","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"ORB-SfMLearner: ORB-Guided Self-supervised Visual Odometry with Selective Online Adaptation\",\"authors\":\"Yanlin Jin, Rui-Yang Ju, Haojun Liu, Yuzhong Zhong\",\"doi\":\"arxiv-2409.11692\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Deep visual odometry, despite extensive research, still faces limitations in\\naccuracy and generalizability that prevent its broader application. To address\\nthese challenges, we propose an Oriented FAST and Rotated BRIEF (ORB)-guided\\nvisual odometry with selective online adaptation named ORB-SfMLearner. We\\npresent a novel use of ORB features for learning-based ego-motion estimation,\\nleading to more robust and accurate results. We also introduce the\\ncross-attention mechanism to enhance the explainability of PoseNet and have\\nrevealed that driving direction of the vehicle can be explained through\\nattention weights, marking a novel exploration in this area. To improve\\ngeneralizability, our selective online adaptation allows the network to rapidly\\nand selectively adjust to the optimal parameters across different domains.\\nExperimental results on KITTI and vKITTI datasets show that our method\\noutperforms previous state-of-the-art deep visual odometry methods in terms of\\nego-motion accuracy and generalizability.\",\"PeriodicalId\":501130,\"journal\":{\"name\":\"arXiv - CS - Computer Vision and Pattern Recognition\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Computer Vision and Pattern Recognition\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.11692\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Computer Vision and Pattern Recognition","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.11692","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
ORB-SfMLearner: ORB-Guided Self-supervised Visual Odometry with Selective Online Adaptation
Deep visual odometry, despite extensive research, still faces limitations in
accuracy and generalizability that prevent its broader application. To address
these challenges, we propose an Oriented FAST and Rotated BRIEF (ORB)-guided
visual odometry with selective online adaptation named ORB-SfMLearner. We
present a novel use of ORB features for learning-based ego-motion estimation,
leading to more robust and accurate results. We also introduce the
cross-attention mechanism to enhance the explainability of PoseNet and have
revealed that driving direction of the vehicle can be explained through
attention weights, marking a novel exploration in this area. To improve
generalizability, our selective online adaptation allows the network to rapidly
and selectively adjust to the optimal parameters across different domains.
Experimental results on KITTI and vKITTI datasets show that our method
outperforms previous state-of-the-art deep visual odometry methods in terms of
ego-motion accuracy and generalizability.