CRAFT: Contextual Re-Activation of Filters for face recognition Training

Simple item page
dc.catalogadorpau
dc.contributor.authorBhatta, Aman
dc.contributor.authorMery, Domingo
dc.contributor.authorWu, Haiyu
dc.contributor.authorBowyer, Kevin W.
dc.date.accessioned2025-03-26T19:02:23Z
dc.date.available2025-03-26T19:02:23Z
dc.date.issued2023
dc.description.abstractThe first layer of a deep CNN backbone applies filters to an image to extract the basic features available to later layers. During training, some filters may go inactive, mean ing all weights in the filter approach zero. An inactive fil ter in the final model represents a missed opportunity to extract a useful feature. This phenomenon is especially prevalent in specialized CNNs such as for face recogni tion (as opposed to, e.g., ImageNet). For example, in one the most widely face recognition model (ArcFace), about half of the convolution filters in the first layer are inactive. We propose a novel approach designed and tested specif ically for face recognition networks, known as "CRAFT: Contextual Re-Activation of Filters for Face Recognition Training". CRAFT identifies inactive filters during training and reinitializes them based on the context of strong filters at that stage in training. We show that CRAFT reduces fraction of inactive filters from 44% to 32% on average and discovers filter patterns not found by standard training. Compared to standard training without reactivation, CRAFT demonstrates enhanced model accuracy on standard face-recognition benchmark datasets including AgeDB-30, CPLFW, LFW, CALFW, and CFP-FP, as well as on more challenging datasets like IJBB and IJBC.
dc.fechaingreso.objetodigital2025-03-26
dc.fuente.origenORCID
dc.identifier.doi10.48550/arXiv.2312.00072
dc.identifier.urihttps://doi.org/10.48550/arXiv.2312.00072
dc.identifier.urihttps://repositorio.uc.cl/handle/11534/103002
dc.information.autorucEscuela de Ingeniería; Mery, Domingo; 0000-0003-4748-3882; 102382
dc.language.isoen
dc.nota.accesocontenido completo
dc.rightsacceso abierto
dc.subject.ddc000
dc.subject.deweyCiencias de la computaciónes_ES
dc.titleCRAFT: Contextual Re-Activation of Filters for face recognition Training
dc.typepreprint
sipa.codpersvinculados102382
sipa.trazabilidadORCID;2025-02-28
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
craft contextual re activation of filters.pdf
Size:
1.02 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Artículos de revistas