{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T14:15:42Z","timestamp":1780668942421,"version":"3.54.1"},"reference-count":39,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2023,3,16]],"date-time":"2023-03-16T00:00:00Z","timestamp":1678924800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Robot. AI"],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Wearable assistive devices for the visually impaired whose technology is based on video camera devices represent a challenge in rapid evolution, where one of the main problems is to find computer vision algorithms that can be implemented in low-cost embedded devices.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Objectives and Methods<\/jats:title>\n                    <jats:p>This work presents a Tiny You Only Look Once architecture for pedestrian detection, which can be implemented in low-cost wearable devices as an alternative for the development of assistive technologies for the visually impaired.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The recall results of the proposed refined model represent an improvement of 71% working with four anchor boxes and 66% with six anchor boxes compared to the original model. The accuracy achieved on the same data set shows an increase of 14% and 25%, respectively. The F1 calculation shows a refinement of 57% and 55%. The average accuracy of the models achieved an improvement of 87% and 99%. The number of correctly detected objects was 3098 and 2892 for four and six anchor boxes, respectively, whose performance is better by 77% and 65% compared to the original, which correctly detected 1743 objects.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>Finally, the model was optimized for the Jetson Nano embedded system, a case study for low-power embedded devices, and in a desktop computer. In both cases, the graphics processing unit (GPU) and central processing unit were tested, and a documented comparison of solutions aimed at serving visually impaired people was performed.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>We performed the desktop tests with a RTX 2070S graphics card, and the image processing took about 2.8\u00a0ms. The Jetson Nano board could process an image in about 110\u00a0ms, offering the opportunity to generate alert notification procedures in support of visually impaired mobility.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frobt.2023.1052509","type":"journal-article","created":{"date-parts":[[2023,3,16]],"date-time":"2023-03-16T03:06:52Z","timestamp":1678936012000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":13,"title":["Pedestrian detection model based on Tiny-Yolov3 architecture for wearable devices to visually impaired assistance"],"prefix":"10.3389","volume":"10","author":[{"given":"Sergio-Uriel","family":"Maya-Mart\u00ednez","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Amadeo-Jos\u00e9","family":"Arg\u00fcelles-Cruz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zobeida-Jezabel","family":"Guzm\u00e1n-Zavaleta","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Miguel-de-Jes\u00fas","family":"Ram\u00edrez-Cadena","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2023,3,16]]},"reference":[{"key":"B1","first-page":"265","article-title":"Tensorflow: A system for large-scale machine learning","author":"Abadi","year":"2016"},{"key":"B2","doi-asserted-by":"publisher","first-page":"2265","DOI":"10.1007\/s11063-020-10197-9","article-title":"An evaluation of retinanet on indoor object detection for blind and visually impaired persons assistance navigation","volume":"51","author":"Afif","year":"2020","journal-title":"Neural Process. Lett."},{"key":"B3","unstructured":"Inicia la \u201cnueva normalidad\u201d y el pa\u00eds est\u00e1 cerca de 10 mil muertos por Covid-192020"},{"key":"B4","doi-asserted-by":"publisher","first-page":"3782","DOI":"10.1109\/TITS.2019.2892405","article-title":"A survey on 3d object detection methods for autonomous driving applications","volume":"20","author":"Arnold","year":"2019","journal-title":"IEEE Trans. Intelligent Transp. Syst."},{"key":"B5","unstructured":"jetson-stats\n            BonghiR.\n          2020"},{"key":"B6","doi-asserted-by":"publisher","first-page":"1483","DOI":"10.1109\/tpami.2019.2956516","article-title":"Cascade R-CNN: High quality object detection and instance segmentation","volume":"43","author":"Cai","year":"2019","journal-title":"IEEE Trans. Pattern Analysis Mach. Intell."},{"key":"B7","doi-asserted-by":"publisher","first-page":"20651","DOI":"10.1007\/s11042-017-5472-5","article-title":"Real-time pedestrian crossing lights detection algorithm for the visually impaired","volume":"77","author":"Cheng","year":"2018","journal-title":"Multimedia Tools Appl."},{"key":"B10","first-page":"5332","article-title":"DeepFashion2: A versatile benchmark for detection, pose estimation, segmentation and re-identification of clothing images","author":"Ge","year":"2019"},{"key":"B12","first-page":"1440","article-title":"Fast R-CNN","author":"Girshick","year":"2015"},{"key":"B11","first-page":"580","article-title":"Rich feature hierarchies for accurate object detection and semantic segmentation","author":"Girshick","year":"2014"},{"key":"B13","volume-title":"MobileNets: Efficient convolutional neural networks for mobile vision applications","author":"Howard","year":"2017"},{"key":"B14","first-page":"2503","article-title":"YOLO-LITE: A real-time object detection algorithm optimized for non-GPU computers","author":"Huang","year":"2018"},{"key":"B15","first-page":"448","article-title":"\u201cBatch normalization: accelerating deep network training by reducing internal covariate shift,\u201d in Proceedings of the 32nd International Conference on Machine Learning, eds. Bach, F., and Blei, D., (Lille, France: PMLR)","author":"Ioffe","year":"2015"},{"key":"B16","doi-asserted-by":"publisher","first-page":"507","DOI":"10.1109\/THMS.2020.3027534","article-title":"An ai-based visual aid with integrated reading assistant for the completely blind","volume":"50","author":"Khan","year":"2020","journal-title":"IEEE Trans. Human-Machine Syst."},{"key":"B17","doi-asserted-by":"publisher","first-page":"1575","DOI":"10.1109\/TIP.2018.2878349","article-title":"A richly annotated pedestrian dataset for person retrieval in real surveillance scenarios","volume":"28","author":"Li","year":"2018","journal-title":"IEEE Trans. image Process."},{"key":"B18","doi-asserted-by":"publisher","first-page":"2476","DOI":"10.3390\/s18082476","article-title":"Visual localizer: Outdoor localization based on convnet descriptor and global optimization for visually impaired pedestrians","volume":"18","author":"Lin","year":"2018","journal-title":"Sensors"},{"key":"B19","doi-asserted-by":"publisher","first-page":"21","DOI":"10.1007\/978-3-319-46448-0_2","article-title":"SSD: Single shot multibox detector","volume":"9905","author":"Liu","year":"2016","journal-title":"Computer Vision\u2014ECCV 2016. ECCV 2016. Lecture Notes in Computer Science"},{"key":"B20","doi-asserted-by":"publisher","first-page":"1089","DOI":"10.1007\/s10462-018-9641-3","article-title":"Recent progress in semantic image segmentation","volume":"52","author":"Liu","year":"2019","journal-title":"Artif. Intell. Rev."},{"key":"B21","doi-asserted-by":"publisher","first-page":"480","DOI":"10.1016\/j.eswa.2017.09.029","article-title":"Abnormal behavior recognition for intelligent video surveillance systems: A review","volume":"91","author":"Mabrouk","year":"2018","journal-title":"Expert Syst. Appl."},{"key":"B22","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1016\/j.aci.2018.01.001","article-title":"A robust single and multiple moving object detection, tracking and classification","volume":"17","author":"Mahalingam","year":"2018","journal-title":"Appl. Comput. Inf."},{"key":"B23","doi-asserted-by":"publisher","first-page":"649","DOI":"10.1109\/TITS.2017.2780621","article-title":"Mechatronic system to help visually impaired users during walking and running","volume":"19","author":"Mancini","year":"2018","journal-title":"IEEE Trans. Intelligent Transp. Syst."},{"key":"B24","doi-asserted-by":"publisher","first-page":"249","DOI":"10.1007\/s10209-020-00764-1","article-title":"User-centered system design for assisted navigation of visually impaired individuals in outdoor cultural environments","volume":"21","author":"Ntakolia","year":"2022","journal-title":"Univ. Access Inf. Soc."},{"key":"B25","volume-title":"TensorRT SDK (Computer software)","year":"2020"},{"key":"B26","first-page":"779","article-title":"You only look once: Unified, real-time object detection","author":"Redmon","year":"2016"},{"key":"B27","volume-title":"YOLOv3: An incremental improvement","author":"Redmon","year":"2018"},{"key":"B28","first-page":"4510","article-title":"Mobilenetv2: Inverted residuals and linear bottlenecks","author":"Sandler","year":"2018"},{"key":"B29","volume-title":"COVID-Robot: Monitoring social distancing constraints in crowded scenarios","author":"Sathyamoorthy","year":"2020"},{"key":"B42","doi-asserted-by":"publisher","first-page":"4537","DOI":"10.3390\/s22124537","article-title":"A wearable assistive device for blind pedestrians using real-time object detection and tactile presentation","volume":"22","author":"Shen","year":"2022","journal-title":"Sensors"},{"key":"B31","doi-asserted-by":"publisher","first-page":"11771","DOI":"10.1007\/s11042-016-3617-6","article-title":"A computer vision-based perception system for visually impaired","volume":"11771","author":"Tapu","year":"2017","journal-title":"Multimed. Tools Appl."},{"key":"B32","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1016\/j.patrec.2018.10.031","article-title":"Wearable assistive devices for visually impaired: A state of the art survey","volume":"137","author":"Tapu","year":"2018","journal-title":"Pattern Recognit. Lett."},{"key":"B33","article-title":"Recent trends in computer vision-driven scene understanding for vi\/blind users: A systematic mapping","volume-title":"Univ. Access Inf. Soc.","author":"Valipoor","year":"2022"},{"key":"B34","volume-title":"Deep learning algorithms with applications to video analytics for A smart city: A survey","author":"Wang","year":"2015"},{"key":"B36","volume-title":"WHO global disability action plan 2014-2021: Better health for all people with disability","year":"2015"},{"key":"B35","volume-title":"World report on vision. Tech. rep","year":"2019"},{"key":"B37","doi-asserted-by":"publisher","first-page":"354","DOI":"10.1016\/j.neucom.2019.01.079","article-title":"Deep learning-based methods for person re-identification: A comprehensive review","volume":"337","author":"Wu","year":"2019","journal-title":"Neurocomputing"},{"key":"B38","first-page":"1369","article-title":"Adversarial examples for semantic segmentation and object detection","author":"Xie","year":"2017"},{"key":"B39","doi-asserted-by":"publisher","first-page":"972","DOI":"10.1109\/JRFID.2022.3212907","article-title":"An efficient pedestrian detection for realtime surveillance systems based on modified yolov3","volume":"6","author":"Xu","year":"2022","journal-title":"IEEE J. Radio Freq. Identif."},{"key":"B40","doi-asserted-by":"publisher","first-page":"4354","DOI":"10.1109\/TIP.2016.2590322","article-title":"Pedestrian behavior modeling from stationary crowds with applications to intelligent surveillance","volume":"25","author":"Yi","year":"2016","journal-title":"IEEE Trans. image Process."},{"key":"B41","first-page":"566","article-title":"Learning data augmentation strategies for object detection","volume-title":"Computer vision \u2013 ECCV 2020. ECCV 2020. Lecture notes in computer science","author":"Zoph","year":"2020"}],"container-title":["Frontiers in Robotics and AI"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frobt.2023.1052509\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,5]],"date-time":"2026-06-05T13:43:14Z","timestamp":1780666994000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frobt.2023.1052509\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,3,16]]},"references-count":39,"alternative-id":["10.3389\/frobt.2023.1052509"],"URL":"https:\/\/doi.org\/10.3389\/frobt.2023.1052509","relation":{},"ISSN":["2296-9144"],"issn-type":[{"value":"2296-9144","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,3,16]]},"article-number":"1052509"}}