Performance of face recognition systems can be adversely affected by mismatches between training and test poses, especially when there is only one training image available. We address this problem by extending each st...
详细信息
Performance of face recognition systems can be adversely affected by mismatches between training and test poses, especially when there is only one training image available. We address this problem by extending each statistical frontal face model with artificially synthesized models for non-frontal views. The synthesis methods are based on several implementations of maximum likelihood linear regression (MLLR), as well as standard multivariate linear regression (LinReg). All synthesis techniques utilize prior information on how face models for the frontal view are related to face models for non-frontal views. The synthesis and extension approach is evaluated by applying it to two face verification systems: PCA based (holistic features) and DCTmod2 based (local features). Experiments on the FERET database suggest that for the PCA based system, the LinReg technique (which is based on a common relation between two sets of points) is more suited than the MLLR based techniques (which in effect are "single point to single point" transforms). For the DCTmod2 based system, the results show that synthesis via a new MLLR implementation obtains better performance than synthesis based on traditional MLLR (due to a lower number of free parameters). The results further show that extending frontal models considerably reduces errors.
In the framework of a face verification system using local features and a Gaussian mixture model based classifier, we address the problem of non-frontal face verification (when only a single (frontal) training image i...
详细信息
In the framework of a face verification system using local features and a Gaussian mixture model based classifier, we address the problem of non-frontal face verification (when only a single (frontal) training image is available) by extending each client's frontal face model with artificially synthesized models for non-frontal views. Furthermore, we propose the maximum likelihood shift (MLS) synthesis technique and compare its performance against a maximum likelihood linear regression (MLLR) based technique (originally developed for adapting speech recognition systems) and the recently proposed "difference between two universal background models" (UBMdiff) technique. All techniques rely on prior information and learn how a generic face model for the frontal view is related to generic models at non-frontal views. Experiments on the FERET database suggest that that the proposed MLS technique is more suitable than MLLR (due to a lower number of free parameters) and UBMdiff (due to lack of heuristics). The results further suggest that extending frontal models considerably reduces errors.
This paper describes recent work conducted in collaboration with DSTO Australia and Aerosonde to develop video-based geolocation (targeting) and real-time enhancement of ground features observed from a small unmanned ...
详细信息
This paper describes recent work conducted in collaboration with DSTO Australia and Aerosonde to develop video-based geolocation (targeting) and real-time enhancement of ground features observed from a small unmanned air vehicle (UAV). Here a combination of low-cost GPS, video and attitude sensors are used to estimate the object ground position whilst a simple super-resolution technique is used to enhance images of the object of interest. The results of two recent trials of the developed algorithms are presented.
Two dimensional electro-optic sampling allows a target's scattered terahertz field to be measured. By obtaining multiple projection images we have performed the reconstruction of three dimensional targets with mil...
详细信息
Two dimensional electro-optic sampling allows a target's scattered terahertz field to be measured. By obtaining multiple projection images we have performed the reconstruction of three dimensional targets with millimeter resolution.
Summary form only given. T-ray computed tomography imaging has significant potential in a number of applications. It is capable of reconstructing the 3D structure and frequency dependent far-infrared optical propertie...
详细信息
Summary form only given. T-ray computed tomography imaging has significant potential in a number of applications. It is capable of reconstructing the 3D structure and frequency dependent far-infrared optical properties of an object. Using the reconstructed material properties different substances may be uniquely identified despite being hidden within other opaque structures. Applications of this technology are foreseen in non-destructive mail/packaging inspection, semiconductor testing, quality control of plastics and certain biomedical applications where the absorption of THz is not prohibitive.
暂无评论