{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,11]],"date-time":"2025-10-11T02:17:51Z","timestamp":1760149071553,"version":"build-2065373602"},"reference-count":53,"publisher":"MDPI AG","issue":"13","license":[{"start":{"date-parts":[[2023,7,4]],"date-time":"2023-07-04T00:00:00Z","timestamp":1688428800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Berlin Center for Digital Transformation","award":["EFRE 1.8\/20"],"award-info":[{"award-number":["EFRE 1.8\/20"]}]},{"name":"Berlin Senate","award":["EFRE 1.8\/20"],"award-info":[{"award-number":["EFRE 1.8\/20"]}]},{"name":"European Union","award":["EFRE 1.8\/20"],"award-info":[{"award-number":["EFRE 1.8\/20"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>We investigate an edge-computing scenario for robot control, where two similar neural networks are running on one computational node. We test the feasibility of using a single object-detection model (YOLOv5) with the benefit of reduced computational resources against the potentially more accurate independent and specialized models. Our results show that using one single convolutional neural network (for object detection and hand-gesture classification) instead of two separate ones can reduce resource usage by almost 50%. For many classes, we observed an increase in accuracy when using the model trained with more labels. For small datasets (a few hundred instances per label), we found that it is advisable to add labels with many instances from another dataset to increase detection accuracy.<\/jats:p>","DOI":"10.3390\/s23136138","type":"journal-article","created":{"date-parts":[[2023,7,5]],"date-time":"2023-07-05T00:53:04Z","timestamp":1688518384000},"page":"6138","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":6,"title":["Data Fusion for Cross-Domain Real-Time Object Detection on the Edge"],"prefix":"10.3390","volume":"23","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8285-5173","authenticated-orcid":false,"given":"Mykyta","family":"Kovalenko","sequence":"first","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, 10587 Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8459-9182","authenticated-orcid":false,"given":"David","family":"Przewozny","sequence":"additional","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, 10587 Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8378-4805","authenticated-orcid":false,"given":"Peter","family":"Eisert","sequence":"additional","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, 10587 Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8104-3904","authenticated-orcid":false,"given":"Sebastian","family":"Bosse","sequence":"additional","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, 10587 Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8807-0008","authenticated-orcid":false,"given":"Paul","family":"Chojecki","sequence":"additional","affiliation":[{"name":"Fraunhofer Heinrich Hertz Institute, 10587 Berlin, Germany"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2023,7,4]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1109\/MIM.2019.8674634","article-title":"Deep learning-based hand gesture recognition for collaborative robots","volume":"22","author":"Nuzzi","year":"2019","journal-title":"IEEE Instrum. Meas. Mag."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"231","DOI":"10.1006\/cviu.1995.1017","article-title":"A Survey of Automated Visual Inspection","volume":"61","author":"Newman","year":"1995","journal-title":"Comput. Vis. Image Underst."},{"key":"ref_3","doi-asserted-by":"crossref","unstructured":"Nath, V., and Chattopadhyay, C. (2021, January 19\u201322). S2D2Net: An Improved Approach For Robust Steel Surface Defects Diagnosis with Small Sample Learning. Proceedings of the 2021 IEEE International Conference on Image Processing (ICIP), Anchorage, AK, USA.","DOI":"10.1109\/ICIP42928.2021.9506405"},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Cambeiro, A., Seibold, C., Hilsmann, A., and Eisert, P. (2022, January 6\u20138). Automated Damage Inspection of Power Transmission Towers from UAV Images. Proceedings of the International Conference on Computer Vision Theory and Applications, Online.","DOI":"10.5220\/0010826500003124"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"661","DOI":"10.1007\/s40684-021-00343-6","article-title":"State of the Art in Defect Detection Based on Machine Vision","volume":"9","author":"Ren","year":"2022","journal-title":"Int. J. Precis. Eng.-Manuf.-Green Technol."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Krishna Chand Gudi, S.L., Ojha, S., Johnston, B., Clark, J., and Williams, M.A. (2018, January 1\u20133). Fog Robotics for Efficient, Fluent and Robust Human-Robot Interaction. Proceedings of the 2018 IEEE 17th International Symposium on Network Computing and Applications (NCA), Cambridge, MA, USA.","DOI":"10.1109\/NCA.2018.8548077"},{"key":"ref_7","unstructured":"Padmanabhan, A., Agarwal, N., Iyer, A., Ananthanarayanan, G., Shu, Y., Karianakis, N., Xu, G.H., and Netravali, R. (2023, January 17\u201319). Gemel: Model Merging for {Memory-Efficient}, {Real-Time} Video Analytics at the Edge. Proceedings of the 20th USENIX Symposium on Networked Systems Design and Implementation, Boston, MA, USA."},{"key":"ref_8","unstructured":"Jeong, J.S., Kim, S., Yu, G.I., Lee, Y., and Chun, B.G. (2020). Accelerating Multi-Model Inference by Merging DNNs of Different Weights. Technical Report. arXiv."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Doll\u00e1r, P., and Zitnick, C.L. (2014, January 6\u201312). Microsoft coco: Common objects in context. Proceedings of the Computer Vision\u2013ECCV 2014: 13th European Conference, Zurich, Switzerland. Proceedings, Part V 13.","DOI":"10.1007\/978-3-319-10602-1_48"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"842","DOI":"10.1109\/COMST.2021.3061435","article-title":"Resource Allocation and Service Provisioning in Multi-Agent Cloud Robotics: A Comprehensive Survey","volume":"23","author":"Afrin","year":"2021","journal-title":"IEEE Commun. Surv. Tutor."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"108177","DOI":"10.1016\/j.comnet.2021.108177","article-title":"Task offloading in Edge and Cloud Computing: A survey on mathematical, artificial intelligence and control theory solutions","volume":"195","author":"Saeik","year":"2021","journal-title":"Comput. Netw."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"27","DOI":"10.1007\/s43154-020-00005-6","article-title":"Review of Interfaces for Industrial Human-Robot Interaction","volume":"1","author":"Berg","year":"2020","journal-title":"Curr. Robot. Rep."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"207635","DOI":"10.1109\/ACCESS.2020.3037724","article-title":"Robots and Wizards: An Investigation Into Natural Human\u2013Robot Interaction","volume":"8","author":"Strazdas","year":"2020","journal-title":"IEEE Access"},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"157422","DOI":"10.1109\/ACCESS.2021.3129650","article-title":"A Review of the Hand Gesture Recognition System: Current Progress and Future Directions","volume":"9","author":"Mohamed","year":"2021","journal-title":"IEEE Access"},{"key":"ref_15","doi-asserted-by":"crossref","unstructured":"Rahman, A., Kovalenko, M., Przewozny, D., Dawoud, K., Chojecki, P., Eisert, P., and Bosse, S. (2021, January 17\u201320). Can You Do Real-Time Gesture Recognition with 5 Watts?. Proceedings of the 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Melbourne, Australia.","DOI":"10.1109\/SMC52423.2021.9659029"},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Jezek, S., Jonak, M., Burget, R., Dvorak, P., and Skotak, M. (2021, January 25\u201327). Deep learning-based defect detection of metal parts: Evaluating current methods in complex conditions. Proceedings of the 2021 13th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT), Brno, Czech Republic.","DOI":"10.1109\/ICUMT54235.2021.9631567"},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Andrei-Alexandru, T., Cosmin, C., Bogdan, P., and Adrian-Alexandru, T. (2021, January 1\u20133). Automated ceramic plate defect detection using ScaledYOLOv4-large. Proceedings of the 2021 13th International Conference on Electronics, Computers and Artificial Intelligence (ECAI), Pitesti, Romania.","DOI":"10.1109\/ECAI52376.2021.9515185"},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"170461","DOI":"10.1109\/ACCESS.2020.3021508","article-title":"Exploring Deep Learning-Based Architecture, Strategies, Applications and Current Trends in Generic Object Detection: A Comprehensive Review","volume":"8","author":"Aziz","year":"2020","journal-title":"IEEE Access"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"3503","DOI":"10.1007\/s12555-020-0118-1","article-title":"Machine Learning-based Automatic Optical Inspection System with Multimodal Optical Image Fusion Network","volume":"19","author":"Lee","year":"2021","journal-title":"Int. J. Control. Autom. Syst."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"105651","DOI":"10.1016\/j.engappai.2022.105651","article-title":"Multi-task learning on the edge for effective gender, age, ethnicity and emotion recognition","volume":"118","author":"Foggia","year":"2023","journal-title":"Eng. Appl. Artif. Intell."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"500","DOI":"10.1111\/mice.12755","article-title":"Finicky transfer learning\u2014A method of pruning convolutional neural networks for cracks classification on edge devices","volume":"37","author":"Miszczak","year":"2022","journal-title":"Comput.-Aided Civ. Infrastruct. Eng."},{"key":"ref_22","unstructured":"(2023, June 15). Pruning\/Sparsity Tutorial\u2014Ultralytics YOLOv8 Docs. Available online: https:\/\/docs.ultralytics.com\/yolov5\/tutorials\/model_pruning_and_sparsity\/#test-normally."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"e6317","DOI":"10.1002\/cpe.6317","article-title":"Performance characterization of video analytics workloads in heterogeneous edge infrastructures","volume":"35","author":"Rivas","year":"2023","journal-title":"Concurr. Comput. Pract. Exp."},{"key":"ref_24","unstructured":"Xiao, Z., Xia, Z., Zheng, H., Zhao, B.Y., and Jiang, J. (2021, January 14\u201317). Towards Performance Clarity of Edge Video Analytics. Proceedings of the 2021 IEEE\/ACM Symposium on Edge Computing (SEC), San Jose, CA, USA."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"1530","DOI":"10.1109\/TCC.2022.3146615","article-title":"Nimbus: Towards Latency-Energy Efficient Task Offloading for AR Services","volume":"11","author":"Cozzolino","year":"2022","journal-title":"IEEE Trans. Cloud Comput."},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"4423","DOI":"10.1109\/LRA.2020.2998414","article-title":"RILaaS: Robot Inference and Learning as a Service","volume":"5","author":"Tanwani","year":"2020","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Hanhirova, J., K\u00e4m\u00e4r\u00e4inen, T., Sepp\u00e4l\u00e4, S., Siekkinen, M., Hirvisalo, V., and Yl\u00e4-J\u00e4\u00e4ski, A. (2018, January 12\u201315). Latency and throughput characterization of convolutional neural networks for mobile computer vision. Proceedings of the 9th ACM Multimedia Systems Conference, Amsterdam, The Netherlands.","DOI":"10.1145\/3204949.3204975"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"139","DOI":"10.1109\/LCA.2020.3023723","article-title":"GPU-NEST: Characterizing Energy Efficiency of Multi-GPU Inference Servers","volume":"19","author":"Jahanshahi","year":"2020","journal-title":"IEEE Comput. Archit. Lett."},{"key":"ref_29","unstructured":"Jain, P., Mo, X., Jain, A., Subbaraj, H., Durrani, R.S., Tumanov, A., Gonzalez, J., and Stoica, I. (2018). Dynamic space-time scheduling for gpu inference. arXiv."},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Srivastava, A., Nguyen, D., Aggarwal, S., Luckow, A., Duffy, E., Kennedy, K., Ziolkowski, M., and Apon, A. (2018, January 10\u201313). Performance and memory trade-offs of deep learning object detection in fast streaming high-definition images. Proceedings of the 2018 IEEE International Conference on Big Data (Big Data), Seattle, WA, USA.","DOI":"10.1109\/BigData.2018.8622389"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Fang, J., Liu, Q., and Li, J. (2021, January 24\u201326). A Deployment Scheme of YOLOv5 with Inference Optimizations Based on the Triton Inference Server. Proceedings of the 2021 IEEE 6th International Conference on Cloud Computing and Big Data Analytics (ICCCBDA), Chengdu, China.","DOI":"10.1109\/ICCCBDA51879.2021.9442557"},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Padmanabhan, A., Iyer, A.P., Ananthanarayanan, G., Shu, Y., Karianakis, N., Xu, G.H., and Netravali, R. (2021, January 25). Towards memory-efficient inference in edge video analytics. Proceedings of the 3rd ACM Workshop on Hot Topics in Video Analytics and Intelligent Edges, New Orleans, LA, USA.","DOI":"10.1145\/3477083.3480150"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Lee, J., Liu, Y., and Lee, Y. (2021, January 24\u201325). ParallelFusion: Towards Maximum Utilization of Mobile GPU for DNN Inference. Proceedings of the 5th International Workshop on Embedded and Mobile Deep Learning, EMDL\u201921, Virtual.","DOI":"10.1145\/3469116.3470014"},{"key":"ref_34","doi-asserted-by":"crossref","first-page":"399","DOI":"10.1016\/j.future.2022.03.039","article-title":"Towards automatic model specialization for edge video analytics","volume":"134","author":"Rivas","year":"2022","journal-title":"Future Gener. Comput. Syst."},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Kim, D., Lee, S., Sung, N.M., and Choe, C. (2023, January 20\u201323). Real-time object detection using a domain-based transfer learning method for resource-constrained edge devices. Proceedings of the 2023 International Conference on Artificial Intelligence in Information and Communication (ICAIIC), Bali, Indonesia.","DOI":"10.1109\/ICAIIC57133.2023.10067064"},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Wang, C.Y., Bochkovskiy, A., and Liao, H.Y.M. (2023, January 18\u201322). YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition, Vancouver, BC, Canada.","DOI":"10.1109\/CVPR52729.2023.00721"},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Stone, T., Stone, N., Jain, P., Jiang, Y., Kim, K.H., and Nelakuditi, S. (2019, January 10\u201313). Towards Scalable Video Analytics at the Edge. Proceedings of the 2019 16th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON), Boston, MA, USA.","DOI":"10.1109\/SAHCN.2019.8824876"},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"2585","DOI":"10.1007\/s10115-021-01605-0","article-title":"Model complexity of deep learning: A survey","volume":"63","author":"Hu","year":"2021","journal-title":"Knowl. Inf. Syst."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"549","DOI":"10.1016\/j.procir.2020.05.095","article-title":"Enabling reliable visual quality control in smart factories through TSN","volume":"88","author":"Popper","year":"2020","journal-title":"Procedia CIRP"},{"key":"ref_40","unstructured":"Vick, A., and Krueger, J. (2018, January 20\u201321). Using OPC UA for distributed industrial robot control. Proceedings of the ISR 2018; 50th International Symposium on Robotics, Munich, Germany."},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Liu, W., Anguelov, D., Erhan, D., Szegedy, C., Reed, S., Fu, C.Y., and Berg, A.C. (2016, January 11\u201314). Ssd: Single shot multibox detector. Proceedings of the Computer Vision\u2013ECCV 2016: 14th European Conference, Amsterdam, The Netherlands. Proceedings, Part I 14.","DOI":"10.1007\/978-3-319-46448-0_2"},{"key":"ref_42","unstructured":"Jocher, G., Stoken, A., Chaurasia, A., Borovec, J., NanoCode012, TaoXie, Kwon, Y., Michael, K., Changyu, L., and Fang, J. (2021). Available online: https:\/\/doi.org\/10.5281\/zenodo.5563715."},{"key":"ref_43","first-page":"91","article-title":"Faster r-cnn: Towards real-time object detection with region proposal networks","volume":"28","author":"Ren","year":"2015","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_44","doi-asserted-by":"crossref","unstructured":"Zhao, S., Zheng, J., Sun, S., and Zhang, L. (2022). An Improved YOLO Algorithm for Fast and Accurate Underwater Object Detection. Symmetry, 14.","DOI":"10.2139\/ssrn.4079287"},{"key":"ref_45","doi-asserted-by":"crossref","unstructured":"Padilla, R., Netto, S.L., and da Silva, E.A.B. (2020, January 1\u20133). A Survey on Performance Metrics for Object-Detection Algorithms. Proceedings of the 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), Niteroi, Brazil.","DOI":"10.1109\/IWSSIP48289.2020.9145130"},{"key":"ref_46","doi-asserted-by":"crossref","unstructured":"Padilla, R., Passos, W.L., Dias, T.L., Netto, S.L., and Da Silva, E.A. (2021). A comparative analysis of object detection metrics with a companion open-source toolkit. Electronics, 10.","DOI":"10.3390\/electronics10030279"},{"key":"ref_47","doi-asserted-by":"crossref","unstructured":"Liu, G., Nouaze, J.C., Touko Mbouembe, P.L., and Kim, J.H. (2020). YOLO-tomato: A robust algorithm for tomato detection based on YOLOv3. Sensors, 20.","DOI":"10.3390\/s20072145"},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Kotz, S., and Johnson, N.L. (1992). Breakthroughs in Statistics: Methodology and Distribution, Springer.","DOI":"10.1007\/978-1-4612-4380-9"},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"1026","DOI":"10.21105\/joss.01026","article-title":"Pingouin: Statistics in Python","volume":"3","author":"Vallat","year":"2018","journal-title":"J. Open Source Softw."},{"key":"ref_50","unstructured":"(2022, October 02). Tips for Best Training Results \u00b7 Ultralytics\/yolov5 Wiki \u00b7 GitHub. Available online: https:\/\/github.com\/ultralytics\/yolov5\/wiki\/Tips-for-Best-Training-Results."},{"key":"ref_51","doi-asserted-by":"crossref","first-page":"104120","DOI":"10.1016\/j.trc.2023.104120","article-title":"An automated driving systems data acquisition and analytics platform","volume":"151","author":"Xia","year":"2023","journal-title":"Transp. Res. Part C Emerg. Technol."},{"key":"ref_52","doi-asserted-by":"crossref","first-page":"1956","DOI":"10.1007\/s11263-020-01316-z","article-title":"The open images dataset v4: Unified image classification, object detection, and visual relationship detection at scale","volume":"128","author":"Kuznetsova","year":"2020","journal-title":"Int. J. Comput. Vis."},{"key":"ref_53","unstructured":"McMahan, B., Moore, E., Ramage, D., Hampson, S., and y Arcas, B.A. (2017, January 20\u201322). Communication-efficient learning of deep networks from decentralized data. Proceedings of the Artificial Intelligence and Statistics, PMLR, Fort Lauderdale, FL, USA."}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/13\/6138\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T20:05:53Z","timestamp":1760126753000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/23\/13\/6138"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,7,4]]},"references-count":53,"journal-issue":{"issue":"13","published-online":{"date-parts":[[2023,7]]}},"alternative-id":["s23136138"],"URL":"https:\/\/doi.org\/10.3390\/s23136138","relation":{},"ISSN":["1424-8220"],"issn-type":[{"type":"electronic","value":"1424-8220"}],"subject":[],"published":{"date-parts":[[2023,7,4]]}}}