Despite rapid advances in face recognition, there remains a clear gap between the performance of still image-based face recognition and video-based face recognition, due to the vast difference in visual quality between the domains and the difficulty of curating diverse large-scale video datasets. This paper addresses both of those challenges, through an image to video feature-level domain adaptation approach, to learn discriminative video frame representations. The framework utilizes large-scale unlabeled video data to reduce the gap between different domains while transferring discriminative knowledge from large-scale labeled still images. Given a face recognition network that is pretrained in the image domain, the adaptation is achieved by (i) distilling knowledge from the network to a video adaptation network through feature matching, (ii) performing feature restoration through synthetic data augmentation and (iii) learning a domain-invariant feature through a domain adversarial discriminator. We further improve performance through a discriminator-guided feature fusion that boosts high-quality frames while eliminating those degraded by video domainspecific factors. Experiments on the YouTube Faces and IJB-A datasets demonstrate that each module contributes to our feature-level domain adaptation framework and substantially improves video face recognition performance to achieve state-of-the-art accuracy. We demonstrate qualitatively that the network learns to suppress diverse artifacts in videos such as pose, illumination or occlusion without being explicitly trained for them.

No abstract

We recently reported that Wuzhi tablet (WZ; Schisandra sphenanthera extract) can inhibit P-glycoprotein (P-gp)-mediated efflux and CYP3A-mediated metabolism of tacrolimus (FK506) and thus increase the blood concentrations of FK506. Major active lignans of WZ include schisandrin A, schisandrin B, schisandrin C, schisandrol A, schisandrol B, and schisantherin A. Whether and how these six lignans affect the pharmacokinetics of FK506 remains unclear. Therefore, this study aimed to investigate the effects of these lignans on the first-pass absorption and metabolism of FK506 and the involved mechanisms in vitro and in vivo. The results showed that whole-blood concentrations of FK506 were increased to different degrees following coadministration of the six lignans, respectively. Schisandrol B showed the strongest effect on the increase of the area under the concentration-time curve, the oral bioavailability, the gut processes affecting availability, and the hepatic availability of FK506. The reduction of intestinal first-pass effect contributed most to the increase in oral bioavailability of FK506 when coadministered with schisandrol B. In vitro transport experiment showed that schisandrin A, schisandrin B, and schisandrol B inhibited P-gp-mediated efflux of FK506. In vitro metabolism study showed that the inhibitory effect of these six lignans on FK506 metabolism was dose-dependent. In conclusion, the exposure of FK506 in rats was increased when coadministered with these lignans, and schisandrol B showed the strongest effect. Lignans of WZ inhibited P-gp-mediated efflux and CYP3A-mediated metabolism of FK506, and the reduction of intestinal first-pass affected by the lignans was the major cause of the increased FK506 oral bioavailability.

No abstract