Learning based automatic face annotation for arbitrary poses and expressions from frontal images only

Statistical approaches for building non-rigid deformable models, such as the Active Appearance Model (AAM), have enjoyed great popularity in recent years, but typically require tedious manual annotation of training images. In this paper, a learning based approach for the automatic annotation of visually deformable objects from a single annotated frontal image is presented and demonstrated on the example of automatically annotating face images that can be used for building AAMs for fitting and tracking. This approach employs the idea of initially learning the correspondences between landmarks in a frontal image and a set of training images with a face in arbitrary poses. Using this learner, virtual images of unseen faces at any arbitrary pose for which the learner was trained can be reconstructed by predicting the new landmark locations and warping the texture from the frontal image. View-based AAMs are then built from the virtual images and used for automatically annotating unseen images, including images of different facial expressions, at any random pose within the maximum range spanned by the virtually reconstructed images. The approach is experimentally validated by automatically annotating face images from three different databases. © 2009 IEEE.

PCB-integrated optical waveguide sensors: An ammonia gas sensor

This paper presents a novel platform for the formation of cost-effective PCB-integrated optical waveguide sensors. The sensor design relies on the use of multimode polymer waveguides that can be formed directly on standard PCBs and commercially-available chemical dyes, enabling the integration of all essential sensor components (electronic, photonic, chemical) on low-cost substrates. Moreover, it enables the detection of multiple analytes from a single device by employing waveguide arrays functionalised with different chemical dyes. The devices can be manufactured with conventional methods of the PCB industry, such as solder-reflow processes and pick-and-place assembly techniques. As a proof of principle, a PCB-integrated ammonia gas sensor is fabricated on a FR4 substrate. The sensor operation relies on the change of the optical transmission characteristics of chemically functionalised optical waveguides in the presence of ammonia molecules. The fabrication and assembly of the sensor unit, as well as fundamental simulation and characterisation studies, are presented. The device achieves a sensitivity of approximately 30 ppm and a linear response up to 600 ppm at room temperature. Finally, the potential to detect multiple analytes from a single device is demonstrated using principal-component analysis. © 1983-2012 IEEE.

A survey of alternative once-through fast reactor core designs

Reprocessing of Light Water Reactor (LWR) spent fuel to recover plutonium or transuranics for use in Sodium cooled Fast Reactors (SFRs) is a distant prospect in the U.S.A. This has motivated our evaluation of potentially cost-effective operation of uranium startup fast reactors (USFRs) in a once-through mode. This review goes beyond findings reported earlier based on a UC fueled MgO reflected SFR to describe a broader parametric study of options. Cores were evaluated for a variety of fuel/coolant/reflector combinations: UC/UZr/UO 2/UN;Na/Pb; MgO/SS/Zr. The challenge is achieving high burnup while minimizing enrichment and respecting both cladding fluence/dpa and reactivity lifetime limits. These parametric studies show that while UC fuel is still the leading contender, UO 2 fuel and ZrH 1.7 moderated metallic fuel are also attractive if UC proves to be otherwise inadequate. Overall, these findings support the conclusion that a competitive fuel cycle cost and uranium utilization compared to LWRs is possible for SFRs operated on a once-through uranium fueled fuel cycle. In addition, eventual transition to TRU recycle mode is studied, as is a small test reactor to demonstrate key features.