Heterogeneous Face Recognition Using Kernel Prototype Similarities

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 35; no. 6; pp. 1410 - 1422
• Main Authors: Klare, Brendan F., Jain, Anil K.
• Format: Journal Article
• Language: English
• Published: Los Alamitos, CA IEEE 01.06.2013; IEEE Computer Society
• Subjects: Applied sciences; Artificial intelligence; Biometric Identification - methods; Computer science; control theory; systems; Data processing. List processing. Character string processing; discriminant analysis; Exact sciences and technology; Face; Face recognition; forensic sketch; Forensics; Heterogeneous face recognition; Humans; Image Enhancement - methods; Image Interpretation, Computer-Assisted - methods; infrared image; Kernel; local descriptors; Mathematics; Memory organisation. Data processing; nonlinear similarity; Pattern recognition. Digital image processing. Computational geometry; Probability and statistics; Probes; Prototypes; random subspaces; Sampling theory, sample surveys; Sciences and techniques of general use; Software; Statistics; thermal image; Training; Prototype; Random sampling; Similarity; Sample size; Image processing; prototypes; local descriptors; Vector space; Image matching; Near infrared spectrum; Imaging; Facies; forensic sketch; random subspaces; Database; Monitoring; Infrared thermography; Computer vision; Discriminant analysis; Probabilistic approach; Forensic science; Face recognition; Image retrieval; infrared image; Computational geometry; Surveillance; thermal image; Heterogeneous face recognition; Thermal imaging; nonlinear similarity
• ISSN: 0162-8828; 1939-3539; 2160-9292; 1939-3539
• DOI: 10.1109/TPAMI.2012.229

Heterogeneous face recognition (HFR) involves matching two face images from alternate imaging modalities, such as an infrared image to a photograph or a sketch to a photograph. Accurate HFR systems are of great value in various applications (e.g., forensics and surveillance), where the gallery databases are populated with photographs (e.g., mug shot or passport photographs) but the probe images are often limited to some alternate modality. A generic HFR framework is proposed in which both probe and gallery images are represented in terms of nonlinear similarities to a collection of prototype face images. The prototype subjects (i.e., the training set) have an image in each modality (probe and gallery), and the similarity of an image is measured against the prototype images from the corresponding modality. The accuracy of this nonlinear prototype representation is improved by projecting the features into a linear discriminant subspace. Random sampling is introduced into the HFR framework to better handle challenges arising from the small sample size problem. The merits of the proposed approach, called prototype random subspace (P-RS), are demonstrated on four different heterogeneous scenarios: 1) near infrared (NIR) to photograph, 2) thermal to photograph, 3) viewed sketch to photograph, and 4) forensic sketch to photograph.