Facial Prior Guided Micro-Expression Generation.

This paper focuses on the facial micro-expression (FME) generation task, which has potential application in enlarging digital FME datasets, thereby alleviating the lack of training data with labels in existing micro-expression datasets. Despite obvious progress in the image animation task, FME generation remains challenging because existing image animation methods can hardly encode subtle and short-term facial motion information. To this end, we present a facial-prior-guided FME generation framework that takes advantage of facial priors for facial motion generation. Specifically, we first estimate the geometric locations of action units (AUs) with detected facial landmarks. We further calculate an adaptive weighted prior (AWP) map, which alleviates the estimation error of AUs while efficiently capturing subtle facial motion patterns. To achieve smooth and realistic synthesis results, we use our proposed facial prior module to guide motion representation and generation modules in mainstream image animation frameworks. Extensive experiments on three benchmark datasets consistently show that our proposed facial prior module can be adopted in image animation frameworks and significantly improve their performance on micro-expression generation. Moreover, we use the generation technique to enlarge existing datasets, thereby improving the performance of general action recognition backbones on the FME recognition task. Our code is available at https://github.com/sysu19351158/FPB-FOMM.

Ultra-stable perovskite quantum dot composites encapsulated with mesoporous SiO2 and PbBr(OH) for white light-emitting diodes.

Lead halide perovskite quantum dots (QDs) with high fluorescence efficiency and high color purity have a broad application prospect in the field of backlight display, but poor stability has been a key factor limiting their commercialization. Herein, we successfully synthesized CsPbBr3 QDs-KIT-6 (CsPbBr3 -K6) composite by using KIT-6 molecular sieve as the limited template with a simple high temperature solid-phase method. Further, the semi-protected CsPbBr3 QDs in KIT-6 frame will spontaneously hydrolyze when encountering water, and finally the double-encapsulated CsPbBr3 QDs-KIT-6@PbBr(OH) (CsPbBr3 -K6@PbBr(OH)) composite are obtained. CsPbBr3 -K6@PbBr(OH) composite shows excellent green emission properties, including a photoluminescence quantum yield (PLQY) (~73%) and a narrow emission linewidth of 25 nm. It is interesting that, the composite has excellent stability, including water stability without attenuation of fluorescence intensity after soaking in water for 60 days, thermal stability of 120°C heating-cooling cycle, and excellent optical stability without attenuation under continuous ultraviolet irradiation.

One-Step Hydrothermal Synthesis of MoO2/MoS2 Nanocomposites as High-Performance Electrode Material for Supercapacitors.

By only changing the ratio of Mo to S source, a distinctive single phase MoO2 or MoS2 and MoO2/MoS2 nanocomposites (NCs) are obtained through a simple one-step hydrothermal method based on CH4N2S as a sulfur source and (NH4)6Mo7O24·4H2O as a source of Mo in oxalic acid. The effect of ratio of Mo to S source on the composition, structure, and electrochemical performance are systematically researched. Due to its unique design, abundant macropores active sites in MoO2/MoS2 NCs induce superior rate property (55.30% capacitance retention to 20 from 1 A g-1) and larger specific capacitance (1667.3 F g-1 at 1 A g-1) and longer cycle life (94.75% after 5000 cycles) as used directly as an electrode. Furthermore, at a power density of 225 W kg-1, a maximal energy density of 21.85 Wh kg-1 is provided by the asymmetric supercapacitor (MoO2/MoS2//AC). The capacitance of asymmetric supercapacitor (ASC) is remarkably enhanced by 129.02% under 5000 cycles at a current density of 1.5 A g-1, demonstrating outstanding cycle property. These results imply the prepared MoO2/MoS2 NCs have promising applications in advanced energy storages. It is important and should be noted that NCs of oxide and sulfide are prepared with only a simple one-step process.

CARes-UNet: Content-aware residual UNet for lesion segmentation of COVID-19 from chest CT images.

PURPOSE: Coronavirus disease 2019 (COVID-19) has caused a serious global health crisis. It has been proven that the deep learning method has great potential to assist doctors in diagnosing COVID-19 by automatically segmenting the lesions in computed tomography (CT) slices. However, there are still several challenges restricting the application of these methods, including high variation in lesion characteristics and low contrast between lesion areas and healthy tissues. Moreover, the lack of high-quality labeled samples and large number of patients lead to the urgency to develop a high accuracy model, which performs well not only under supervision but also with semi-supervised methods. METHODS: We propose a content-aware lung infection segmentation deep residual network (content-aware residual UNet (CARes-UNet)) to segment the lesion areas of COVID-19 from the chest CT slices. In our CARes-UNet, the residual connection was used in the convolutional block, which alleviated the degradation problem during the training. Then, the content-aware upsampling modules were introduced to improve the performance of the model while reducing the computation cost. Moreover, to achieve faster convergence, an advanced optimizer named Ranger was utilized to update the model's parameters during training. Finally, we employed a semi-supervised segmentation framework to deal with the problem of lacking pixel-level labeled data. RESULTS: We evaluated our approach using three public datasets with multiple metrics and compared its performance to several models. Our method outperforms other models in multiple indicators, for instance in terms of Dice coefficient on COVID-SemiSeg Dataset, CARes-UNet got the score 0.731, and semi-CARes-UNet further boosted it to 0.776. More ablation studies were done and validated the effectiveness of each key component of our proposed model. CONCLUSIONS: Compared with the existing neural network methods applied to the COVID-19 lesion segmentation tasks, our CARes-UNet can gain more accurate segmentation results, and semi-CARes-UNet can further improve it using semi-supervised learning methods while presenting a possible way to solve the problem of lack of high-quality annotated samples. Our CARes-UNet and semi-CARes-UNet can be used in artificial intelligence-empowered computer-aided diagnosis system to improve diagnostic accuracy in this ongoing COVID-19 pandemic.

In-air self-reduction synthesis and luminescence properties from Mg21 Ca4 Na4 (PO4 )18 :Eu2+ -Eu3+ excited using ultraviolet light.

A series of Mg21 Ca4 Na4 (PO4 )18 :Eu2+ -Eu3+ phosphors was successfully synthesized using a high-temperature solid-state method in an air atmosphere. The phase structures and luminescence properties of the samples were studied in detail. The phosphors exhibited strong visible light emission under different wavelengths of ultraviolet light excitation. By harmonizing the doping concentration of Eu3+ to adjust the relative luminescence intensity of Eu2+ and Eu3+ , a colourful emission of phosphors could be achieved. In addition, the colour coordinates of the International Commission on lighting indicated that Mg21 Ca4 Na4 (PO)18 :Eu2+ -Eu3+ could be considered as a potential blue, orange and red phosphor for white light-emitting diode applications.