Search

Photocatalysis is an advanced oxidation process where light exposure triggers a semiconducting nanomaterial (nano-photocatalyst) to generate electron-hole (e−/h+) pairs and free radicals. This phenomenon is widely used for the photocatalysis-assisted removal of organic and other contaminants using wide range of nano-photocatalysts, offering an efficient approach to environmental remediation. However, the introduction of powdered nano-photocatalysts into water systems often leads to unintended secondary pollution in the form of residual nano-photocatalysts, ion leaching, free radicals, toxic by-products etc. Such practices potentially introduce emerging secondary contaminants into aquatic environments, posing risks to both aquatic life and human health. The resulting chemical by-products and intermediates can effectively induce chronic toxicity, neurological and developmental disorders, cardiovascular defects, and intestinal ailments in humans and aquatic species. Despite having a range of health and environmental consequences, this dark side of nano-photocatalysts has been comparatively less explored and discussed in the literature. In this review, the pros and cons of powder nano-photocatalysts are discussed in view of their advantages as well as disadvantages in wastewater treatment. The discussion encompasses their classification based on composition, dimensions, structure, and activity, as well as recent advancements in improving their photocatalytic efficiency. The article also explores the recent advances on their applications in photocatalytic removal of various water pollutants/contaminants of emerging concern (i.e., organic pollutants, micro/nano plastics, heavy ions, disinfections, etc.) Furthermore, an emphasis on the role of such nano-photocatalysts as emerging (secondary) contaminants in water system, along with a thorough discussion of latest studies related to the health and environmental issues, has been discussed. Additionally, it addresses critical issues in applying powder nano-photocatalysts for wastewater detoxification and explores potential solutions to these challenges followed by future prospects.

In the field of face frontalization, the model obtained by training on a particular dataset often underperforms on other datasets. This paper presents the Pre-trained Feature Transformation GAN (PFT-GAN), which is designed to fully utilize diverse facial feature information available from pre-trained face recognition networks. For that purpose, we propose the use of the feature attention transformation (FAT) module that effectively transfers the low-level facial features to the facial generator. On the other hand, in the hope of reducing the pre-trained encoder dependency, we attempt a new FAT module organization that accommodates the features from all pre-trained face recognition networks employed. This paper attempts evaluating the proposed work using the “independent critic” as well as “dependent critic”, which enables objective judgments. Experimental results show that the proposed method significantly improves the face frontalization performance and helps overcome the bias associated with each pre-trained face recognition network employed.

Abstract The shear-rolling process is a promising directed self-assembly method that can produce high-quality sub−10 nm block copolymer line-space patterns cost-effectively and straightforwardly over a large area. This study presents a high temperature (280 °C) and rapid (~0.1 s) shear-rolling process that can achieve a high degree of orientation in a single process while effectively preventing film delamination, that can be applied to large-area continuous processes. By minimizing adhesion, normal forces, and ultimate shear strain of the polydimethylsiloxane pad, shearing was successfully performed without peeling up to 280 °C at which the chain mobility significantly increases. This method can be utilized for various high-χ block copolymers and surface neutralization processes. It enables the creation of block copolymer patterns with a half-pitch as small as 8 nm in a unidirectional way. Moreover, the 0.1-second rapid shear-rolling was successfully performed on long, 3-inch width polyimide flexible films to validate its potential for the roll-to-roll process.

People are expected to have more opportunities to spend their free time inside the vehicle with advanced vehicle automation in the near future. This will enable people to turn their attention to desirable activities other than driving and to have varied in-vehicle interactions through multimodal ways of conveying and receiving information. Previous studies on in-vehicle multimodal interactions primarily have focused on making users evaluate the impacts of particular multimodal integrations on them, which do not fully provide an overall understanding of user expectations of the multimodal experience in autonomous vehicles. The research was thus designed to fill the research gap by posing the key question “What are the critical aspects that differentiate and characterise in-vehicle multimodal experiences?” To answer this question, five sessions of design fiction workshops were separately conducted with 17 people to understand the users’ expectations of the multimodal experience in autonomous vehicles. Twenty-two subthemes of users’ expected tasks of multimodal experience were extracted through thematic analysis. The research found that two dimensions, attention and duration, are critical aspects that impact in-vehicle multimodal interactions. With this knowledge, a conceptual model of the users’ in-vehicle multimodal experience was proposed with a two-dimensional spectrum, which populates four different layers: sustained, distinct, concurrent, and coherent. The proposed conceptual model could help designers understand and approach users’ expectations more clearly, allowing them to make more informed decisions from the initial stages of the design process.

This paper presents an effective method to boost the RSSI, the received signal strength on a moving path of LoRa wireless networking without change in digital circuitry or power amplifiers. Replacing the wire antenna as the conventionally used sensor by a printed Yagi-Uda antenna as a novel one designed to have a higher gain signifies increments of the RSSI at distant positions in the LoRa system with the same battery. The electromagnetic properties of the novel antenna are given by observing its beam-patterns in the anechoic chamber following the design procedures and fabrication. And the field-test for the radio link is conducted to see the proposed antenna enhance the RSSI at positions along the street of a wide-open area for LoRa communication. The RSSI has risen by more than 10 dB at distances from 200m to 700m with DLP-RFS1280 as the LoRa device in comparison to the conventional configuration in the device.

To train deep learning models for vision-based action recognition of elders’ daily activities, we need large-scale activity datasets acquired under various daily living environments and conditions. However, most public datasets used in human action recognition either differ from or have limited coverage of elders’ activities in many aspects, making it challenging to recognize elders’ daily activities well by only utilizing existing datasets. Recently, such limitations of available datasets have actively been compensated by generating synthetic data from realistic simulation environments and using those data to train deep learning models. In this paper, based on these ideas we develop ElderSim, an action simulation platform that can generate synthetic data on elders’ daily activities. For 55 kinds of frequent daily activities of the elders, ElderSim generates realistic motions of synthetic characters with various adjustable data-generating options and provides different output modalities including RGB videos, two- and three-dimensional skeleton trajectories. We then generate KIST SynADL, a large-scale synthetic dataset of elders’ activities of daily living, from ElderSim and use the data in addition to real datasets to train three state-of-the-art human action recognition models. From the experiments following several newly proposed scenarios that assume different real and synthetic dataset configurations for training, we observe a noticeable performance improvement by augmenting our synthetic data. We also offer guidance with insights for the effective utilization of synthetic data to help recognize elders’ daily activities.

We present a robust skeleton-based action recognition method with graph convolutional network (GCN) that uses the new adjacency matrix, called Rank-GCN. In Rank-GCN, the biggest change from previous approaches is how the adjacency matrix is generated to accumulate features from neighboring nodes by re-defining “adjacency.” The new adjacency matrix, which we call the rank adjacency matrix, is generated by ranking all the nodes according to metrics including the Euclidean distance from the nodes of interest, whereas the previous GCNs methods used only 1-hop neighboring nodes to construct adjacency. By adopting the rank adjacency matrix, we find not only performance improvements but also robustness against swapping, location shifting and dropping of certain nodes. The fact that the human-made rank adjacency matrix wins against the deep-learning-based matrix, implies that there are still some parts that need touch of humans. We expect our Rank-GCN can make performance improvements especially when the predicted human joints are less accurate and unstable.

The whole world is struggling with current coronavirus pandemic that shows urgent need to develop novel technologies, medical innovations or innovative materials for controlling SARS-CoV-2 infection. The mode of infection of SARS-CoV-2 is still not well known and seems to spread through surface, air, and water. Therefore, the whole surrounding environment needs to be disinfected with continuous function. For that purpose, materials with excellent antiviral properties, cost effective, environmental friendly and practically applicable should be researched. Titanium dioxide (TiO2) under ultraviolet light produces strong oxidative effect and is utilized as photocatalytic disinfectant in biomedical field. TiO2 based photocatalysts are effective antimicrobial/antiviral agents under ambient conditions with potential to be used even in indoor environment for inactivation of bacteria/viruses. Interestingly, recent studies highlight the effective disinfection of SARS-CoV-2 using TiO2 photocatalysts. Here, scope of TiO2 photocatalysts as emerging disinfectant against SARS-CoV-2 infection has been discussed in view of their excellent antibacterial and antiviral activities against various bacteria and viruses (e.g. H1N1, MNV, HSV, NDV, HCoV etc.). The current state of development of TiO2 based nano-photocatalysts as disinfectant shows their potential to combat with SARS-CoV-2 viral infection and are promising for any other such variants or viruses, bacteria in future studies.

This study proposes a novel face detector called DEFace that focuses on the challenging tasks of face detection to cope with a small size that is under 12 pixels and occlusions due to a mask or human body parts. This study proposed the extended feature pyramid network (FPN) module to detect small faces by expanding the range of P layer, and the network by adding a receptive context module (RCM) after each predicted feature head from the top-down pathway in the FPN architecture to enhance the feature discriminability and the robustness. Based on the FPN principle, the combination between the low- and high-resolutions are beneficial for object detection with different object sizes. Furthermore, with assistance from the RCM, the proposed method can use a broad range of context information especially for small faces. To evaluate the performance of the proposed method, various public face datasets are used such as the WIDER Face dataset, the face detection dataset and benchmark (FDDB), and the masked faces (MAFA) dataset, which consist of challenging samples such as small face regions and occlusions by hair or other people. The results indicate that DEFace can detect the face region more accurately in comparison to the other state-of-the-art methods while maintaining the processing time.

Global and intercontinental networking relies on satellite communication. Its wireless communication system always has antennas and their feed assembly comprising waveguides. This makes the satellite payload heavy and costly. In this paper, a novel method is proposed to effectively reduce the size of a waveguide bandpass filter (BPF). Because the metallic cavities make the conventional waveguide end up with a large geometry, especially for high-order BPFs, very compact waveguide-type resonators having metamaterial zeroth-order resonance (WG-ZOR) are designed on the cross-section of the waveguide and substituted for the cavities. While the cavities are half-wavelength resonators, the WG-ZOR is shorter than one eighth of a wavelength. A substantial reduction in size and weight of the waveguide filter is observed as the resonators are cascaded in series through coupling elements in the X-band much longer than K- or Ka-band. An X-band of 7.25~7.75 GHz is chosen to verify the method as the passband with attenuation of 40 dB at 7.00 GHz and 8.00 GHz as the roll-off in the stopband. The BPF is manufactured using the CNC milling technique. The design is carried out with geometrical parameters, not of the level of 10 μm, but the level of 100 μm, which is good for manufacturers but a big challenge for component designers. The measurement of the manufactured metal waveguide filter reveals that the passband has about ≤1 dB and ≤−15 dB as insertion loss and reflection coefficient and the stopband has ≤−40 dB as attenuation, which are in good agreement with the results of the circuit and simulation. The proposed filter has a length of 3.5 λg as the eighth-order BPF, but the conventional waveguide is 5 λg as the seventh-order BPF for the same area of the cross-section. This metamaterial BPF is combined with a horn antenna. The filter enables the wide-band antenna to distinguish the band of transmission from that of noise suppression. This channel selectivity is obviously observed by the filter integrated antenna test.

In this paper, a novel method is proposed to effectively reduce the size of a waveguide bandpass filter (BPF). Because the metallic cavities make the conventional waveguide end up with a large geometry, especially for high-order BPFs, very compact waveguide-type resonators having metamaterial zeroth-order resonance (WG ZOR) are designed on the cross section of the waveguide and substituted for the cavities. While the cavities are half-wavelength resonators, the WG ZOR is shorter than one-eighth of a wavelength. A substantial reduction in the size and weight of the waveguide filter is observed as the resonators are cascaded in series through coupling elements in the X-band that is much longer than that in K- or Ka-bands. The proposed metamaterial filter is realized as a 3D-printed structure to be lighter and thus more suitable for low earth orbit (LEO) satellites. An X-band of 7.25–7.75 GHz is chosen to verify the method as the passband with an attenuation of 40 dB at 7.00 GHz and 8.00 GHz as the roll-off in the stopband. The BPF is manufactured in two ways, namely the CNC-milling technique and metal coating–added 3D printing. The design is carried out with a geometrical parameter of not 10−2 mm but rather 10-1 mm, which is good for manufacturers but challenging for component designers. The measurement of the manufactured metal waveguide filters reveals that the passband has about ≤1 dB and ≤−15 dB as the insertion loss and the reflection coefficient, respectively, and the stopband has an attenuation of ≤−40 dB, which are in good agreement with the results of the circuit and the simulation. The proposed filter has a length of 14 cm as the eighth-order BPF, but the conventional waveguide is 20 cm as the seventh-order BPF for the same area of the cross section.

In this paper, a novel thin and flexible antenna is proposed for earbuds to gain an improvement in their wireless signal-sensing capability as a film-based artificial magnetic conductor (AMC) structure. As antenna designs for earbuds face challenges of being embedded beneath the top cover of the earbud, conformal to curved surfaces, and very close to metallic ground and touch-panel parts, as well as scarce degrees of freedom from feeding conditions and functional degradation by human tissue, unlike conventional techniques such as quasi quarter-wavelength radiators on LDS and epoxy molding compounds (relatively thick and pricy), an antenna of a metal pattern on a film is made with another film layer as the AMC to mitigate problems of the antenna in a small and curved space of an insert-molded wireless device. The antenna was designed, fabricated, and embedded in earbud mockups to work for the 2.4 GHz Bluetooth RF link, and its functions were verified by RF and antenna measurement, showing that it could overcome the limitations in impedance matching with only lumped elements and poor radiation by the ordinary schemes. The input reflection coefficient and antenna efficiency were 10 dB and 9% better than other methods. In particular, the on-film AMC antenna (OFAA) presents robustness against deterioration by the human tissue, when it is placed in the ear phantom at the workbench and implemented in an in situ test using a large zorb ball mimicking a realistic sensing environment. This yielded an RSSI enhancement of 20–30 dB.

In this paper, a novel chip antenna and its function in wireless connectivity are presented for Bluetooth (BLT) earphones. The chip antenna is a metamaterial so compact (3, that when it is mounted on the realistic PCB, it can be held in the enclosure of the BLT earphone. This setting does not degrade the resonance (S11 < −10 dB) of the proposed antenna. As two earphones in a pair are demanded to communicate with each other, one shares an RF signal with the other and they take turns as the master and slave. The received signal sensing is conducted with the latest model of human head-ear-phantom located between the earphones to mimic the real use-case and cross-head interference. Electromagnetic simulation of the antenna is done and verified by fabrication and measurement. Particularly, received-signal strength indications between the proposed antennas in the earphones are experimentally obtained as −67.5 dBm and −70 dBm without and with the head-ear-phantom, respectively, much greater than −120 dBm, the limit of detection, and implying acceptable connectivity and invulnerability over cross-head-interference problems.

Custom inspection using X-ray imaging is a very promising application of modern pattern recognition technology. However, the lack of data or renewal of tariff items makes the application of such technology difficult. In this paper, we present a data augmentation technique based on a new image-to-image translation method to deal with these difficulties. Unlike the conventional methods that convert a semantic label image into a realistic image, the proposed method takes a texture map with a special modification as an additional input of a generative adversarial network to reproduce domain-specific characteristics, such as background clutter or sensor-specific noise patterns. The proposed method was validated by applying it to backscatter X-ray (BSX) vehicle data augmentation. The Fréchet inception distance (FID) of the result indicates the visual quality of the translated image was significantly improved from the baseline when the texture parameters were used. Additionally, in terms of data augmentation, the experimental results of classification, segmentation, and detection show that the use of the translated image data, along with the real data consistently, improved the performance of the trained models. Our findings show that detailed depiction of the texture in translated images is crucial for data augmentation. Considering the comparatively few studies that have examined custom inspections of container scale goods, such as cars, we believe that this study will facilitate research on the automation of container screening, and the security of aviation and ports.

In this paper, an intuitive approach to assessing advantages of beamforming in 5G wireless communication is proposed as a novel try and practical demonstration of importance of alignment between the transmitter’s and receiver’s beams working in millimeter-wave frequency bands. Since the diffraction loss of millimeter-wave signals matters seriously in propagation, the effects of the misalignment and alignment between beams need to be checked for, which was conducted with a horn antenna and the 4 × 4 Butler matrix which mimic the relationship of the base station and handset antennas. Designing and using the microstrip-line and the substrate integrated waveguide (SIW) Butler matrices, RF-to-RF wireless connectivity between the horn and the microstrip line beamformer as case 1 and the horn and the SIW beamformer as case 2, concerning the changing angle of the beam from either of the two Butler matrices, was tested, showing over 12 dB enhancement in received power. This direct electromagnetic link test was accompanied by examining 64-QAM constellations for beam-angle changing from −30° to +30° for the two cases, where the error vector magnitude in the QAM-diagram becomes less than 10% by beam-alignment for the changing angle.

Facial expressions are one of the important non-verbal ways used to understand human emotions during communication. Thus, acquiring and reproducing facial expressions is helpful in analyzing human emotional states. However, owing to complex and subtle facial muscle movements, facial expression modeling from images with face poses is difficult to achieve. To handle this issue, we present a method for acquiring facial expressions from a non-frontal single photograph using a 3D-aided approach. In addition, we propose a contour-fitting method that improves the modeling accuracy by automatically rearranging 3D contour landmarks corresponding to fixed 2D image landmarks. The acquired facial expression input can be parametrically manipulated to create various facial expressions through a blendshape or expression transfer based on the FACS (Facial Action Coding System). To achieve a realistic facial expression synthesis, we propose an exemplar-texture wrinkle synthesis method that extracts and synthesizes appropriate expression wrinkles according to the target expression. To do so, we constructed a wrinkle table of various facial expressions from 400 people. As one of the applications, we proved that the expression-pose synthesis method is suitable for expression-invariant face recognition through a quantitative evaluation, and showed the effectiveness based on a qualitative evaluation. We expect our system to be a benefit to various fields such as face recognition, HCI, and data augmentation for deep learning.

Face recognition is one research area that has benefited from the recent popularity of deep learning, namely the convolutional neural network (CNN) model. Nevertheless, the recognition performance is still compromised by the model’s dependency on the scale of input images and the limited number of feature maps in each layer of the network. To circumvent these issues, we propose PSI-CNN, a generic pyramid-based scale-invariant CNN architecture which additionally extracts untrained feature maps across multiple image resolutions, thereby allowing the network to learn scale-independent information and improving the recognition performance on low resolution images. Experimental results on the LFW dataset and our own CCTV database show PSI-CNN consistently outperforming the widely-adopted VGG face model in terms of face matching accuracy.

In this paper, we present a novel framework for automatically assessing facial attractiveness that considers four ratio feature sets as objective elements of facial attractiveness. In our framework, these feature sets are combined with three regression-based predictors to estimate a facial beauty score. To enhance the system’s performance to make it comparable with human scoring, we apply a score fusion technique. Experimental results show that the attractiveness score obtained by the proposed framework better correlates with human assessments than the scores from other predictors. The framework’s modularity allows any features or predictors to be integrated into the facial attractiveness measure. Our proposed framework can be applied to many beauty-related fields, such as the plastic surgery, cosmetics, and entertainment industries.