![Victor Romero Cano](https://profiles-cardiff.imgix.net/attachments/5855?w=200&h=200&auto=format&crop=faces&fit=crop&q=90")
Dr Victor Romero Cano
(e/fe)
Darlithydd
Yr Ysgol Cyfrifiadureg a Gwybodeg
- Ar gael fel goruchwyliwr ôl-raddedig
Trosolwyg
Rwy'n Ddarlithydd yn yr Ysgol Cyfrifiadureg a Gwybodeg, o fewn yr uned ymchwil cyfrifiadura sy'n canolbwyntio ar bobl .
Wedi'i eni yn Palmira, Colombia, rwyf wedi gweithio mewn roboteg a Dysgu Peiriant (ML) ers dros ddegawd. Astudiais beirianneg mecatroneg yn Universidad Autónoma de Occidente (UAO) yn Cali Colombia. Yn 2008, symudais i Brasil, lle dilynais MSc mewn peirianneg drydanol gyda ffocws ar beirianneg systemau yn Ysgol Polytechnig Prifysgol São Paulo. Ar gyfer traethawd hir fy meistr, ac o dan oruchwyliaeth Oswaldo Luiz do Valle Costa, datblygais algorithmau ar gyfer creu cynrychioliadau digidol o amgylcheddau coed-boblog o ddata laser 2D a gasglwyd gan dîm o robotiaid symudol. Yn 2011, symudais i Awstralia ac ymunais â Chanolfan Roboteg Awstralia a Chanolfan Awtomeiddio Mwynglawdd Rio Tinto (RTCMA) ym Mhrifysgol Sydney, lle dilynais fy astudiaethau doethurol diolch i ysgoloriaeth a ddarparwyd gan y cawr mwyngloddio Rio Tinto. Yn ystod fy Ph.D. ac o dan oruchwyliaeth Juan I Nieto a Gabriel Agamennoni, datblygais dechnolegau ar gyfer canfod, olrhain a dosbarthu gwrthrychau mewn amgylcheddau deinamig cymhleth trwy fanteisio ar natur aml-foddol data stereo-weledol.
Yn 2015, ar ôl cwblhau fy PhD, fe wnes i barhau â'm cysylltiad â RTCMA i ddatblygu technegau System Gwybodaeth Ddaearyddol (GIS) a ML ar gyfer dadansoddi awtomatig a darganfod patrymau yn nhalaith weithredol a thraffig mwyngloddiau Rio Tinto. Yn 2016, symudais i Ffrainc i weithio fel ymchwilydd ôl-ddoethurol, o fewn y Robot Navigation Robot Cydweithredol a Dynol-ymwybodol mewn Amgylcheddau Dynamig Tîm (CHROMA) ar brosiect ar y cyd rhwng Toyota Motor Europe ac Inria Grenoble, dan oruchwyliaeth Christian Laugier. Nod fy nghyfraniadau oedd datblygu technolegau adnabod gwrthrychau Camera-LiDAR ar gyfer cymwysiadau gyrru trefol. Gweithiais ar asio data delwedd a laser ar gyfer gwell adnabyddiaeth gwrthrychau yng nghyd-destun canfyddiad trefol o lwyfannau symudol, gan ysgogi gwaith mapio grid meddiannaeth y tîm, ynghyd â dulliau PGM a Dysgu Dwfn. Rwyf hefyd wedi cydweithio â'r tîm ar brosiectau sy'n canolbwyntio ar wneud penderfyniadau.
Yn 2017 ymunais â'r Gyfadran Peirianneg Universidad Autónoma de Occidente (UAO), Colombia, fel roboteg ac athro dysgu peirianydd. Wynebais yr heriau o adeiladu labordy roboteg o'r gwaelod i fyny, a dylunio roboteg gyfoes a modiwlau canfyddiad roboteg a oedd yn caniatáu i fyfyrwyr peirianneg ddatblygu llwybr gyrfa gyda phwyslais ar agweddau cyfrifiadurol roboteg. O ganlyniad, yn ystod fy pum mlynedd yn UAO, arweiniais sefydlu labordy Roboteg a Systemau Ymreolaethol y brifysgol. Yn 2019, diolch i Weinyddiaeth Amaeth a Choedwigaeth Gweriniaeth Twrci, treuliais fis yn Izmir, Twrci fel Gwyddonydd Gwahoddedig yn y Sefydliad Ymchwil Olewydd yn gweithio ar ffenoteipio planhigion gan ddefnyddio gweledigaeth gyfrifiadurol. Yn 2022 a 2023, gweithiais yn yr Adran Ymchwil a Datblygu Gyrru Ymreolaethol yn Technoleg Rimac, Cwmni seiliedig Croatia sy'n dylunio, peirianwyr, ac yn cynhyrchu cydrannau a systemau cerbydau trydan perfformiad uchel. Yn fy rôl fel rheolwr AI / Canfyddiad, arweiniais dîm yn gweithio ar ddysgu peiriant wedi'i fewnosod a thechnegau canfyddiad aml-synhwyrydd ar gyfer monitro amgylchedd a gyrwyr. Cydweithiodd yn agos hefyd ag ymchwilwyr rheoli cynnig a ffactorau dynol.
Cyhoeddiad
2024
- Ferro-Sánchez, C. A., Díaz-Laverde, C. O., Romero Cano, V., Campo, O. and González-Vargas, A. M. 2024. Machine learning for the classification of surgical patients in orthodontics. Presented at: IX Latin American Congress on Biomedical Engineering and XXVIII Brazilian Congress on Biomedical Engineering, Florianópolis, Brazil, 24-28 October 2022IX Latin American Congress on Biomedical Engineering and XXVIII Brazilian Congress on Biomedical Engineering. CLAIB CBEB 2022 2022. IFMBE Proceedings, vol 99, Vol. 99. Cham: Springer pp. 207-217., (10.1007/978-3-031-49404-8_21)
2023
- Romero Cano, V., Vignard, N. and Laugier, C. 2023. Electronic device, system and method for augmenting image data of a passive optical sensor. US11774554B2 [Patent].
- Rivero-Ortega, J. D., Mosquera-Maturana, J. S., Pardo-Cabrera, J., Hurtado-López, J., Hernández, J. D., Romero Cano, V. and Ramírez-Moreno, D. F. 2023. Ring attractor bio-inspired neural network for social robot navigation. Frontiers in Neurorobotics 17, article number: 1211570. (10.3389/fnbot.2023.1211570)
- Obando-Ceron, J. S., Romero Cano, V. and Monteiro, S. 2023. Probabilistic multi-modal depth estimation based on camera–LiDAR sensor fusion. Machine Vision and Applications 34(5), article number: 79. (10.1007/s00138-023-01426-x)
2022
- Gomez-Ayalde, D. and Romero Cano, V. A. 2022. Local planning methods for autonomous navigation on sidewalks: a comparative survey. Presented at: ColCACI 2022: Colombian Conference on Applications of Computational Intelligence, 27-29 July 20222022 IEEE Colombian Conference on Applications of Computational Intelligence (ColCACI). IEEE pp. 1-6., (10.1109/ColCACI56938.2022.9905339)
- Sierra Gonzalez, D., Vignard, N., Laugier, C., Dibangoye, J. and Romero Cano, V. 2022. Driving assistance method and system. US11364899B2 [Patent].
- Saavedra-Ruiz, M., Pinto-Vargas, A. M. and Romero Cano, V. 2022. Monocular visual autonomous landing system for quadcopter drones using software in the loop. IEEE Aerospace and Electronic Systems Magazine 37(5), pp. 2-16. (10.1109/MAES.2021.3115208)
- Tamayo-Monsalve, M. A. et al. 2022. Coffee maturity classification using convolutional neural networks and transfer learning. IEEE Access 10, pp. 42971-42982. (10.1109/ACCESS.2022.3166515)
2021
- Salazar, G., Saavedra, M. and Romero Cano, V. 2021. High-level camera-LiDAR fusion for 3D object detection with machine learning. [Online]. ArXiv: Cornell University. (10.48550/arXiv.2105.11060) Available at: https://doi.org/10.48550/arXiv.2105.11060
2017
- Romero Cano, V., Vignard, N. and Laugier, C. 2017. XDvision: Dense outdoor perception for autonomous vehicles. Presented at: 28th IEEE Intelligent Vehicles Symposium, 11-14 June 20172017 IEEE Intelligent Vehicles Symposium (IV). IEEE pp. 752-757., (10.1109/IVS.2017.7995807)
2015
- Romero-Cano, V., Agamennoni, G. and Nieto, J. 2015. A variational approach to simultaneous multi-object tracking and classification. International Journal of Robotics Research 35(6), pp. 654-671. (10.1177/0278364915583881)
- Romero Cano, V. 2015. Simultaneous multi-object tracking and classification via approximate variational inference. PhD Thesis, The University of Sydney.
2014
- Romero Cano, V., Agamennoni, G. and Nieto, J. 2014. A variational approach to simultaneous tracking and classification of multiple objects. Presented at: 17th International Conference on Information Fusion, Salamanca, 7-10th July 201417th International Conference on Information Fusion (FUSION). IEEE pp. 1-8.
Cynadleddau
- Ferro-Sánchez, C. A., Díaz-Laverde, C. O., Romero Cano, V., Campo, O. and González-Vargas, A. M. 2024. Machine learning for the classification of surgical patients in orthodontics. Presented at: IX Latin American Congress on Biomedical Engineering and XXVIII Brazilian Congress on Biomedical Engineering, Florianópolis, Brazil, 24-28 October 2022IX Latin American Congress on Biomedical Engineering and XXVIII Brazilian Congress on Biomedical Engineering. CLAIB CBEB 2022 2022. IFMBE Proceedings, vol 99, Vol. 99. Cham: Springer pp. 207-217., (10.1007/978-3-031-49404-8_21)
- Gomez-Ayalde, D. and Romero Cano, V. A. 2022. Local planning methods for autonomous navigation on sidewalks: a comparative survey. Presented at: ColCACI 2022: Colombian Conference on Applications of Computational Intelligence, 27-29 July 20222022 IEEE Colombian Conference on Applications of Computational Intelligence (ColCACI). IEEE pp. 1-6., (10.1109/ColCACI56938.2022.9905339)
- Romero Cano, V., Vignard, N. and Laugier, C. 2017. XDvision: Dense outdoor perception for autonomous vehicles. Presented at: 28th IEEE Intelligent Vehicles Symposium, 11-14 June 20172017 IEEE Intelligent Vehicles Symposium (IV). IEEE pp. 752-757., (10.1109/IVS.2017.7995807)
- Romero Cano, V., Agamennoni, G. and Nieto, J. 2014. A variational approach to simultaneous tracking and classification of multiple objects. Presented at: 17th International Conference on Information Fusion, Salamanca, 7-10th July 201417th International Conference on Information Fusion (FUSION). IEEE pp. 1-8.
Erthyglau
- Rivero-Ortega, J. D., Mosquera-Maturana, J. S., Pardo-Cabrera, J., Hurtado-López, J., Hernández, J. D., Romero Cano, V. and Ramírez-Moreno, D. F. 2023. Ring attractor bio-inspired neural network for social robot navigation. Frontiers in Neurorobotics 17, article number: 1211570. (10.3389/fnbot.2023.1211570)
- Obando-Ceron, J. S., Romero Cano, V. and Monteiro, S. 2023. Probabilistic multi-modal depth estimation based on camera–LiDAR sensor fusion. Machine Vision and Applications 34(5), article number: 79. (10.1007/s00138-023-01426-x)
- Saavedra-Ruiz, M., Pinto-Vargas, A. M. and Romero Cano, V. 2022. Monocular visual autonomous landing system for quadcopter drones using software in the loop. IEEE Aerospace and Electronic Systems Magazine 37(5), pp. 2-16. (10.1109/MAES.2021.3115208)
- Tamayo-Monsalve, M. A. et al. 2022. Coffee maturity classification using convolutional neural networks and transfer learning. IEEE Access 10, pp. 42971-42982. (10.1109/ACCESS.2022.3166515)
- Romero-Cano, V., Agamennoni, G. and Nieto, J. 2015. A variational approach to simultaneous multi-object tracking and classification. International Journal of Robotics Research 35(6), pp. 654-671. (10.1177/0278364915583881)
Gosodiad
- Romero Cano, V. 2015. Simultaneous multi-object tracking and classification via approximate variational inference. PhD Thesis, The University of Sydney.
Gwefannau
- Salazar, G., Saavedra, M. and Romero Cano, V. 2021. High-level camera-LiDAR fusion for 3D object detection with machine learning. [Online]. ArXiv: Cornell University. (10.48550/arXiv.2105.11060) Available at: https://doi.org/10.48550/arXiv.2105.11060
Patentau
- Romero Cano, V., Vignard, N. and Laugier, C. 2023. Electronic device, system and method for augmenting image data of a passive optical sensor. US11774554B2 [Patent].
- Sierra Gonzalez, D., Vignard, N., Laugier, C., Dibangoye, J. and Romero Cano, V. 2022. Driving assistance method and system. US11364899B2 [Patent].
Meysydd goruchwyliaeth
Mae gen i ddiddordeb mewn goruchwylio myfyrwyr PhD ym meysydd:
- Canfyddiad robot a gwneud penderfyniadau
- Dysgu o'r dechrau i'r diwedd ar gyfer systemau robotig ac ymreolaethol
- Rhagfynegiad cynnig cypledig a chynllunio robot.
- Ymasiad aml-synhwyrydd ar gyfer roboteg a chymwysiadau synhwyro o bell
- Roboteg maes ac amaethyddol
- Goblygiadau moesegol a chymdeithasol roboteg
- Gyrru ymreolaethol
Themâu ymchwil
Arbenigeddau
- Roboteg
- Deallusrwydd artiffisial