Although many hypo-pigmenting agents are currently available, the demand for novel whitening agents is increasing, in part due to the weak effectiveness and unwanted side effects of currently available compounds. To screen for novel hypo-pigmenting agents, many methodologies such as cell culture and enzymatic assays are routinely used. However, these models have disadvantages in terms of physiological and economic relevance. In this study, we validated zebrafish as a whole-animal model for phenotype-based screening of melanogenic inhibitors or stimulators. We used both the well-known melanogenic inhibitors (1-phenyl-2-thiourea, arbutin, kojic acid, 2-mercaptobenzothiazole) and newly developed small molecule compounds (haginin, YT16i). All the tested compounds produced inhibitory effects on the pigmentation of zebrafish, most likely due to their inhibitory potential on tyrosinase activity. In simultaneous in vivo toxicity tests, a newly developed melanogenic inhibitor YT16i showed massive abnormalities in terms of deformed morphologies and cardiac function. Together, these results provide a rationale in screening and evaluating the putative melanogenic regulatory compounds. We suggest that the zebrafish system is a novel alternative to mammalian models, with several advantages including the rapidity, cost-effectiveness, and physiological relevance.

Collagenase, a matrix metalloproteinases (MMPs), is a key regulator in the photoaging process of skin due to the reactive oxygen species generated after exposure to ultraviolet A (UVA). Flavonoid compounds have been demonstrated to possess antioxidant properties, and could be useful in the prevention of photoaging. In this study, to investigate the structure-activity relationship of flavonoid compounds on their antioxidant property and inhibitory effects against the MMP activity, the effects of several flavonoids; myricetin, quercetin, kaempferol, luteolin, apigenin and chrysin, on the reactive oxygen species scavengering activity and inhibitory effect against the MMP activity were examined in vitro and in human dermal fibroblasts induced by UVA. The relative order of antioxidative efficacy, as determined using the 1, 1-diphenyl-2-picrylhydrazyl (DPPH) method and the xanthine/xanthine oxidase system, was as follows; flavones: luteolin > apigenin > chrysin, flavonols: myricetin > quercetin > kaempferol, and correlated with the respective number of OH group on their B-ring. In good correlation with the antioxidant properties, the flavonoids inhibited the collagenase activities, in a dose-dependent manner, and the MMP expression. These results suggested the UVA induced antioxidative activity and inhibitory effects of flavonoids on the collagenase in human dermal fibroblasts depends on the number of OH group in the flavonoid structure, and those with a higher number of OH group may be more useful in the prevention of UV stressed skin aging.

The paper discusses recent results on the development of localized arc filament plasma actuators and their use in controlling high-speed and high Reynolds number jet flows. Multiple plasma actuators (up to 8) are controlled using a custom-built 8-channel high-voltage pulsed plasma generator. The plasma generator independently controls pulse repetition rate (0-200 kHz), duty cycle and phase for each individual actuator. Current and voltage measurements demonstrated the power consumption of each actuator to be quite low (20 W at 20% duty cycle). Emission spectroscopy temperature measurements in the pulsed arc filament showed rapid temperature increase over the first 10-20 µs of arc operation, from below 1000 • C to up to about 2000 • C. At longer discharge pulse durations, 20-100 µs, the plasma temperature levels off at approximately 2000 • C. Modelling calculations using an unsteady, quasi-one-dimensional arc filament model showed that rapid localized heating in the arc filament on a microsecond time scale generates strong compression waves. The results of the calculations also suggest that flow forcing is most efficient at low actuator duty cycles, with short heating periods and sufficiently long delays between the pulses to allow for convective cooling of high-temperature filaments. The model predictions are consistent with laser sheet scattering flow visualization results and particle imaging velocimetry measurements. These measurements show large-scale coherent structure formation and considerable mixing enhancement in an ideally expanded Mach 1.3 jet forced by eight repetitively pulsed plasma actuators. The effects of forcing are most significant near the jet preferred mode frequency (ν = 5 kHz). The results also show a substantial reduction in the jet potential core length and a significant increase in the jet Mach number decay rate beyond the end of potential core, especially at low actuator duty cycles.

Localized arc filament plasma actuators were used to control an axisymmetric Mach 0.9 jet with a Reynolds number based on the nozzle exit diameter of 7:6 10 5. Eight actuators, distributed azimuthally inside the nozzle, near the nozzle exit, were used to excite various instabilities and azimuthal modes of the jet over a large Strouhal number range (St DF of 0.1 to 5.0). Time-resolved pressure measurements were used to investigate the development of actuation perturbations in the jet, particle image velocimetry measurements were used to evaluate the control effects on the turbulence field, and far-field sound was measured to evaluate the control effects on the radiated acoustic field. The jet responded to the forcing over a large range of excitation Strouhal numbers with varying degrees. As expected, the low Strouhal number seeded disturbances grew slowly, saturated farther downstream, and stayed saturated for a longer time before decaying gradually. The saturation and decay of the seeded perturbations moved farther upstream as their Strouhal number was increased. Seeded perturbations with higher azimuthal modes exhibited faster decay. Particle image velocimetry results showed that when exciting the jet's preferred-mode instability at lower azimuthal modes, the jet potential core was shortened and the turbulent kinetic energy was increased significantly. At higher Strouhal numbers and higher azimuthal modes, forcing had less of an impact on the mean velocity and turbulent kinetic energy. Far-field acoustic results showed a significant noise increase (2 to 4 dB) when the jet is excited around the jet's preferred-mode instability Strouhal number (St DF 0:2-0:5), in agreement with the results in the literature. Noise reduction of 0.5 to over 1.0 dB is observed over a large excitation Strouhal number range; this reduction seems to peak around St DF 1:5 to 2.0 at a 30-deg angle, but around St DF 3:0 to 3.5 at a 90deg angle. Although forcing the jet with higher azimuthal modes is advantageous for noise mitigation at a 30-deg angle and lower Strouhal numbers, the effect is not as clear at higher forcing Strouhal numbers and at a 90-deg angle.

Although electrochemical double-layer capacitors (EDLC) have the advantages of high rate capability, good durability, and superior cycle life over presently available secondary batteries, they have drawbacks in terms of energy density and cost. Especially, their relatively low energy storage capacity is mainly responsible for their limited usage. [ 1 ] To overcome this limitation, various types of hybrid capacitors have been developed and successfully commercialized. The lithium ion capacitor (LIC), one of the hybrid capacitors under development, has a specifi c energy and energy density of around 30 Whkg − 1 and 20 WhL − 1 , respectively. These values are more than three times higher than the values of a conventional EDLC. [ 2 ] LICs consist of an activated carbon cathode (positive electrode) and a carbonaceous lithium-intercalating anode (negative electrode). The fl at discharge profi le of the negative electrode enables improvement of LICs by allowing higher utilization of the cathode within an expanded voltage window. In fabrication, lithium doping of the anode is necessary process prior to charge the capacitor. During the doping process, metallic lithium is oxidized to release lithium ion, while the reduction of the lithium ion occurs at the surface of the anode. Throughout the process, the voltage of the anode decreases until it is close to 0 V (vs. Li/Li + ). The lithium doping is a necessary process for LICs and metallic lithium is currently used as the lithium source. The metallic lithium has to be handled with caution to minimize risks since lithium metal reacts violently with moisture, causing fi re and explosion. [ 3 ] We therefore propose a new lithium-doping method that uses a stable lithium metal oxide, Li 2 MoO 3 as an alternative lithium source, thanks to its huge charge capacity and the irreversibility of the fi rst cycle. [ 4 ] The innovative use of Li 2 MoO 3 simplifi es the lithium-doping process and meets the stringent safety requirements for use of LICs. Herein, the advantages of our novel concept will be discussed in detail.Schematic diagrams comparing different lithium-doping methods are given in Figure 1 . According to the cell confi guration of a conventional LIC, metallic lithium is incorporated in the cell, as illustrated in Figure 1 a. By means of electrical short-circuiting, lithium ions can be inserted into anode materials. In this way, it would be diffi cult to load the amount of desired lithium precisely during cell fabrication because the loaded metallic lithium cannot be fully inserted into the anode materials. Moreover, any remaining metallic lithium in the LIC would cause safety issues. In contrast, our proposed concept for LICs is that Li 2 MoO 3 is integrated into the cathode, as presented in Figure 1 b. Lithium ion can be electrochemically extracted from the Li 2 MoO 3 and then effectively inserted into the anode. The inserted lithium ions are only partially recovered to the Li 2 MoO 3 within the operating-voltage range of LICs. After lithium ion extraction, the delithiate...

Attention networks in multimodal learning provide an efficient way to utilize given visual information selectively. However, the computational cost to learn attention distributions for every pair of multimodal input channels is prohibitively expensive. To solve this problem, co-attention builds two separate attention distributions for each modality neglecting the interaction between multimodal inputs. In this paper, we propose bilinear attention networks (BAN) that find bilinear attention distributions to utilize given vision-language information seamlessly. BAN considers bilinear interactions among two groups of input channels, while low-rank bilinear pooling extracts the joint representations for each pair of channels. Furthermore, we propose a variant of multimodal residual networks to exploit eight-attention maps of the BAN efficiently. We quantitatively and qualitatively evaluate our model on visual question answering (VQA 2.0) and Flickr30k Entities datasets, showing that BAN significantly outperforms previous methods and achieves new state-of-the-arts on both datasets.

In this study, we investigated the expression and putative role of Sox9 in epidermal keratinocyte. Immunohistochemical staining showed that Sox9 is predominantly expressed in the basal layer of normal human skin epidermis, and highly expressed in several skin diseases including psoriasis, basal cell carcinoma, keratoacanthoma and squamous cell carcinoma. In calcium-induced keratinocyte differentiation model, the expression of Sox9 was decreased in a time dependent manner. When Sox9 was overexpressed using a recombinant adenovirus, cell growth was enhanced, while the expression of differentiation-related genes such as loricrin and involucrin was markedly decreased. Similarly, when rat skin was intradermally injected with the adenovirus expressing Sox9, the epidermis was thickened with increase of PCNA positive cells, while the epidermal differentiation was decreased. Finally, UVB irradiation induced Sox9 expression in cultured human epidermal keratinocytes, and keratinocytes are protected from UVB-induced apoptosis by Sox9 overexpression. Together, these results suggest that Sox9 is an important regulator of epidermal keratinocytes with putative pro-proliferation and/or pro-survival functions, and may be related to several cutaneous diseases that are characterized by abnormal differentiation and hyperproliferation.

0000−0003−2426−2198] , Seong-Ho Choi 2[0000−0002−7553−6761] , Jin-Hwa Kim 3 [0000−0002−0423−0415] , and Byoung-Tak Zhang 2,4[0000−0001−9890−0389]Abstract. We propose a video story question-answering (QA) architecture, Multimodal Dual Attention Memory (MDAM). The key idea is to use a dual attention mechanism with late fusion. MDAM uses selfattention to learn the latent concepts in scene frames and captions. Given a question, MDAM uses the second attention over these latent concepts.Multimodal fusion is performed after the dual attention processes (late fusion). Using this processing pipeline, MDAM learns to infer a high-level vision-language joint representation from an abstraction of the full video content. We evaluate MDAM on PororoQA and MovieQA datasets which have large-scale QA annotations on cartoon videos and movies, respectively. For both datasets, MDAM achieves new state-of-the-art results with significant margins compared to the runner-up models. We confirm the best performance of the dual attention mechanism combined with late fusion by ablation studies. We also perform qualitative analysis by visualizing the inference mechanisms of MDAM.