PhD topic
Multiple Object Tracking (MOT)
Following multiple objects through video sequences while keeping each one's identity consistent over time.
Manuel Bendaña
PhD Student · Centro Singular de Investigación en Tecnoloxías Intelixentes (CiTIUS) · Universidade de Santiago de Compostela
Santiago de Compostela, Spain
I am a PhD student at Universidade de Santiago de Compostela and a researcher at Centro Singular de Investigación en Tecnoloxías Intelixentes (CiTIUS), supervised by Prof. Manuel Mucientes and Prof. Víctor M. Brea.
My research focuses on Computer Vision, more specifically on Multiple Object Tracking (MOT), which is the core topic on my PhD. Before starting my PhD, I received an MSc in Computer Vision from Universidade de Santiago de Compostela. Moreover, I have received a highly competitive Formación de Profesorado Universitario (FPU) grant from the Spanish Ministry of Universities to support my PhD studies.
Research interests
My main research interest is deep learning applied to Computer Vision, with a focus on Multiple Object Tracking (MOT). However, I also have experience in related topics such as object detection and object segmentation.
Object detection
Locating and classifying objects in images and video. Tracking-by-Detection (TbD) approaches for MOT rely on object detection as a first step to find the objects to track.
Object segmentation
Delineating the exact shape of each object at pixel level. Segmentation can be helpful to better locate objects in complex scenarios.
Deep learning
Neural network models underlie all my work, from detection and segmentation architectures to the feature extraction methods that describe each object's appearance.
Selected publications
Segmentation-guided association refinement in multiple object tracking
Neurocomputing, vol. 708, pp. 135176, 2027 Available online
@article{BENDANA2027135176,
author = {Manuel Bendaña and Víctor M. Brea and Manuel Mucientes}
title = {Segmentation-guided association refinement in multiple object tracking},
journal = {Neurocomputing},
volume = {708},
pages = {135176},
year = {2027},
issn = {0925-2312},
doi = {10.1016/j.neucom.2026.135176},
}Enhancing Multi-object Tracking with Segmentation Masks: A Solution for Lost Object Recovery
12th Iberian Conference on Pattern Recognition and Image Analysis, 2025 Oral presentation
@inproceedings{BendanaVBM25,
author = {Manuel Benda{\~{n}}a and
Lorenzo Vaquero and
V{\'{\i}}ctor M. Brea and
Manuel Mucientes},
title = {Enhancing Multi-object Tracking with Segmentation Masks: {A} Solution
for Lost Object Recovery},
booktitle = {Pattern Recognition and Image Analysis - 12th Iberian Conference,
IbPRIA 2025, Coimbra, Portugal, June 30 - July 3, 2025, Proceedings,
Part {I}},
series = {Lecture Notes in Computer Science},
volume = {15937},
pages = {68--79},
publisher = {Springer},
year = {2025},
doi = {10.1007/978-3-031-99565-1\_6}
}Projects
The project addresses the development of an object detection system with RGB and thermal camera to be run on an FPGA on board a UAV for two use cases: (i) search, location and rescue of people on land; (ii) fire prevention and firefighting.
Civil UAVs Initiative: Fase B, Xunta de Galicia (Galicia, Spain) - Axencia Galega de Innovación (GAIN)
REOPEN: Leveraging Artificial Intelligence for Robust Predictive Monitoring in Process Mining
The REOPEN project aims to improve predictive models for complex process mining by using generative models to address data issues like imbalances and anomalies. It also develops novel object detection and tracking models for video to extract activities from unstructured processes, validating them in domains like forest fires and logistics.
Ministerio de Ciencia e Innovación (Spain)
Education & experience
Education
-
2024 – Present
PhD in Information Technology Research
Universidade de Santiago de Compostela
Thesis: Visual Multi Object Tracking with frequent occlusions. Advisors: Prof. Manuel Mucientes and Prof. Víctor M. Brea.
-
2022 – 2024
MSc in Computer Vision
-
2018 – 2022
BSc in Computer Science
Experience
-
January 2024 – Present
Predoctoral Researcher
CiTIUS — Centro Singular de Investigación en Tecnoloxías Intelixentes
FPU-funded PhD research on ‘Visual Multi Object Tracking with frequent occlusions’.
-
January 2023 – October 2023
Junior Researcher
CiTIUS — Centro Singular de Investigación en Tecnoloxías Intelixentes
Research on Object Detection and Visual Object Tracking.
-
March 2022 – May 2022
Research Internship
CiTIUS — Centro Singular de Investigación en Tecnoloxías Intelixentes
Curricular Internship (BSc Degree in Computer Science): ‘Development of video analytics systems using detection and tracking techniques’.
-
2021 – 2022
Research Intern
CiTIUS — Centro Singular de Investigación en Tecnoloxías Intelixentes
Research on ‘Object detection using few-shot techniques’, funded by a Collaboration Grant (Spanish Ministry of Education).
-
July 2021
Research Intern
CiTIUS — Centro Singular de Investigación en Tecnoloxías Intelixentes
Summer internship on ‘Object detection in video with few examples’.
Talks & teaching
Talks & posters
-
Talk
Segmentation-Guided Association Refinement in Multiple Object Tracking
Doctoral Meeting · CiTIUS, Santiago de Compostela, Spain
-
Talk
Enhancing Multi-Object Tracking with Segmentation Masks: A Solution for Lost Object Recovery
IbPRIA 2025 — 12th Iberian Conference on Pattern Recognition and Image Analysis · Coimbra, Portugal
Teaching
-
2024-2025
Automata Theory and Formal Languages (BSc in Computer Science)
Teaching assistant · lab sessions · 20 h · Universidade de Santiago de Compostela
-
2024-2025
Neural Networks and Deep Learning (BSc in Artificial Intelligence)
Teaching assistant · lab sessions · 40 h · Universidade de Santiago de Compostela
-
2025-2026
Automata and Formal Languages (BSc in Artificial Intelligence)
Teaching assistant · lab sessions · 60 h · Universidade de Santiago de Compostela
-
2026-2027
Automata Theory and Formal Languages (BSc in Computer Science)
Teaching assistant · lab sessions · 45 h · Universidade de Santiago de Compostela
-
2026-2027
Neural Networks and Deep Learning (BSc in Artificial Intelligence)
Teaching assistant · lab sessions · 15 h · Universidade de Santiago de Compostela
Awards & grants
Awards
-
2025
MSc in Computer Vision: Best Thesis Award, Extraordinary Award and Top Academic Record
Escola Técnica Superior de Enxeñaría, Universidade de Santiago de Compostela
-
2023
Extraordinary Graduation Award (BSc in Computer Science)
Xunta de Galicia, Spain
-
2023
BSc in Computer Science: Extraordinary Award and Top Academic Record
Escola Técnica Superior de Enxeñaría, Universidade de Santiago de Compostela
-
2020 & 2021
Academic Excellence Awards (academic years 2018/2019 and 2019/2020)
Xunta de Galicia, Spain
Grants
-
2023
Formación del Profesorado Universitario (FPU) Predoctoral Research Grant
Ministry of Universities, Spain — Highly competitive national call
-
2021
Collaboration Grant
Ministry of Education, Spain