Visual Multi-Object Tracking through Deep Learning
Por favor, use este identificador para citas ou ligazóns a este ítem:
http://hdl.handle.net/10347/30934
Ficheiros no ítem
Metadatos do ítem
Título: | Visual Multi-Object Tracking through Deep Learning |
Autor/a: | Vaquero Otal, Lorenzo |
Dirección/Titoría: | Mucientes Molina, Manuel Brea Sánchez, Víctor Manuel |
Centro/Departamento: | Universidade de Santiago de Compostela. Escola de Doutoramento Internacional (EDIUS) Universidade de Santiago de Compostela. Programa de Doutoramento en Investigación en Tecnoloxías da Información |
Palabras chave: | tracking | motion estimation | deep learning | computer vision | |
Data: | 2023 |
Resumo: | This thesis presents novel deep-learning approaches for tracking multiple simultaneous objects in videos, which is an essential component in several applications such as robotics or video surveillance. Traditional multi-object tracking systems, which rely on frame-by-frame detections and primarily geometric attributes, are ill-suited for real-time environments and open-set scenarios. To overcome these limitations, we introduce SiamMT, an innovative architecture that adapts single-object tracking techniques for handling multiple arbitrary targets in real time. This approach is further refined with SiamMOTION, which effectively manages distractors and accommodates objects of varying sizes by extracting semantically-richer features and proposing more accurate search areas. Lastly, this thesis proposes a Transformer-based architecture that complements a lightweight detector by recovering undetected objects for enhanced performance in multiple object tracking. |
Data de Embargo: | 2024-07-24 |
URI: | http://hdl.handle.net/10347/30934 |
Dereitos: | Attribution-NonCommercial-NoDerivatives 4.0 Internacional |
Coleccións
O ítem ten asociados os seguintes ficheiros de licenza: