A2-LINK: Recognizing Disguised Faces via Active Learning and Adversarial Noise based Inter-Domain Knowledge

IEEE Transactions on Biometrics, Behavior, and Identity Science 2020 Anshuman SuriMayank VatsaRicha Singh

Face recognition in the unconstrained environment is an ongoing research challenge. Although several covariates of face recognition such as pose and low resolution have received significant attention“, disguise” is considered an onerous covariate of face recognition... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK BENCHMARK
Heterogeneous Face Recognition Disguised Faces in the Wild A2-LINK GAR @0.1% FAR Impersonation 69.27 # 2
GAR @0.1% FAR Obfuscation 93.08 # 1
GAR @0.1% FAR Overall 93.01 # 1
GAR @1% FAR Impersonation 99.01 # 1
GAR @1% FAR Obfuscation 95.93 # 1
GAR @1% FAR Overall 95.99 # 1
Heterogeneous Face Recognition Disguised Faces in the Wild 2019 A-LINK GAR @0.1% FAR Impersonation 76.40 # 2
GAR @0.1% FAR Obfuscation 96.84 # 2
GAR @0.1% FAR Plastic Surgery 95.20 # 2
GAR @0.1% FAR Overall 95.96 # 2
GAR @0.01% FAR Impersonation 52.80 # 2
GAR @0.01% FAR Obfuscation 94.02 # 2
GAR @0.01% FAR Plastic Surgery 92.00 # 2
GAR @0.01% FAR Overall 93.06 # 2
Heterogeneous Face Recognition Disguised Faces in the Wild 2019 A2-LINK GAR @0.1% FAR Impersonation 79.2 # 1
GAR @0.1% FAR Obfuscation 99.00 # 1
GAR @0.1% FAR Plastic Surgery 98.80 # 1
GAR @0.1% FAR Overall 98.63 # 1
GAR @0.01% FAR Impersonation 54.40 # 1
GAR @0.01% FAR Obfuscation 97.20 # 1
GAR @0.01% FAR Plastic Surgery 96.00 # 1
GAR @0.01% FAR Overall 96.18 # 1

Methods used in the Paper