Learning Soft Mask Based Feature Fusion with Channel and Spatial Attention for Robust Visual Object Tracking

We propose to improve the visual object tracking by introducing a soft mask based low-level feature fusion technique. The proposed technique is further strengthened by integrating channel and spatial attention mechanisms. The proposed approach is integrated within a Siamese framework to demonstrate...

Descripción completa

Detalles Bibliográficos
Autores principales: Fiaz, Mustansar, Mahmood, Arif, Jung, Soon Ki
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2020
Materias:
Article
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7412361/
https://www.ncbi.nlm.nih.gov/pubmed/32698339
http://dx.doi.org/10.3390/s20144021

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7412361/
https://www.ncbi.nlm.nih.gov/pubmed/32698339
http://dx.doi.org/10.3390/s20144021

Ejemplares similares