Face detection based on a human attention guided multi-scale model
Heuristics
ENCODE
Pyramid (geometry)
Representation
DOI:
10.1007/s00422-023-00978-5
Publication Date:
2023-12-01T13:02:18Z
AUTHORS (3)
ABSTRACT
Abstract Multiscale models are among the cutting-edge technologies used for face detection and recognition. An example is Deformable part-based (DPMs), which encode a as multiplicity of local areas (parts) at different resolution scales their hierarchical spatial relationship. Although these have proven successful incredibly efficient in practical applications, mutual position parts involved arbitrarily defined by human specialist final choice optimal based on heuristics. This work seeks to understand whether multi-scale model can take inspiration from fixations select specific scales. In more detail, it shows that pyramid representation be adopted extract interesting points, attention points lead best performance. Human therefore provide valid methodological basis build multiscale model, selecting interest most relevant humans.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES (39)
CITATIONS (3)
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....