Facile hydrothermal assembly of three-dimensional GO-MTZE composite and its adsorption properties toward Cu2.

Three-dimensional (3D) graphene oxide (GO)-based aerogels, GO and 4-methyl-5-thiazoleethanol (MTZE) composites, were prepared by a facile hydrothermal method. Due to the hydrogen bonding and π-π stacking interactions, the produced 3D GO-MTZE composites possessed large cylindrical structures. The morphologies, composition, and chemical states of 3D GO-MTZE3:1 composite were characterized by Fourier transform infrared (FT-IR) spectroscopy, X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), energy-dispersive X-ray spectroscopy (EDS), and N2 adsorption-desorption isotherms based on the Brunauer-Emmett-Teller (BET) method. The existence of nitrogen (N)-containing heterocyclic system and oxygen (O)-containing branched chain of MTZE contributed to the formation of 3D structures, while the complexation effect of heterocyclic sulfur (S)- and N-containing functional groups of MTZE for metal cations dominated the adsorption performance of 3D GO-MTZE3:1 composite, which could selectively adsorb copper ions (Cu2+). In addition, the better hydrophobic property of 3D GO-MTZE3:1 composite facilitates its facile recycling from aqueous solution after adsorption. The adsorption data of 3D GO-MTZE3:1 composite toward Cu2+ fitted well (R2 = 0.9996) with the linear pseudo-second-order kinetic model, giving an equilibrium rate constant (k2) of 0.0187 g mg-1 min-1. The linear Langmuir isothermal model could more accurately describe the experimental data, indicating the adsorption process is mainly dominated by the complexation interactions between MTZE and Cu2+. The thermodynamic parameters of ΔG° (< 0), ΔH° (> 0), and ΔS° (> 0) further indicate that the adsorption is a spontaneous and endothermic, confirming that the complexation between Cu2+ and 3D GO-MTZE3:1 composite occurs. Due to its high selectivity for Cu2+, good hydrophobicity, and excellent stability, the developed 3D GO-MTZE3:1 composite possesses might be promisingly used in the aqueous selective enrichment/removal of Cu2+.

Learning to Holistically Detect Bridges From Large-Size VHR Remote Sensing Imagery.

Bridge detection in remote sensing images (RSIs) plays a crucial role in various applications, but it poses unique challenges compared to the detection of other objects. In RSIs, bridges exhibit considerable variations in terms of their spatial scales and aspect ratios. Therefore, to ensure the visibility and integrity of bridges, it is essential to perform holistic bridge detection in large-size very-high-resolution (VHR) RSIs. However, the lack of datasets with large-size VHR RSIs limits the deep learning algorithms' performance on bridge detection. Due to the limitation of GPU memory in tackling large-size images, deep learning-based object detection methods commonly adopt the cropping strategy, which inevitably results in label fragmentation and discontinuous prediction. To ameliorate the scarcity of datasets, this paper proposes a large-scale dataset named GLH-Bridge comprising 6,000 VHR RSIs sampled from diverse geographic locations across the globe. These images encompass a wide range of sizes, varying from 2,048 × 2,048 to 16,384 × 16,384 pixels, and collectively feature 59,737 bridges. These bridges span diverse backgrounds, and each of them has been manually annotated, using both an oriented bounding box (OBB) and a horizontal bounding box (HBB). Furthermore, we present an efficient network for holistic bridge detection (HBD-Net) in large-size RSIs. The HBD-Net presents a separate detector-based feature fusion (SDFF) architecture and is optimized via a shape-sensitive sample re-weighting (SSRW) strategy. The SDFF architecture performs inter-layer feature fusion (IFF) to incorporate multi-scale context in the dynamic image pyramid (DIP) of the large-size image, and the SSRW strategy is employed to ensure an equitable balance in the regression weight of bridges with various aspect ratios. Based on the proposed GLH-Bridge dataset, we establish a bridge detection benchmark including the OBB and HBB tasks, and validate the effectiveness of the proposed HBD-Net. Additionally, cross-dataset generalization experiments on two publicly available datasets illustrate the strong generalization capability of the GLH-Bridge dataset. The dataset and source code will be released at https://luo-z13.github.io/GLH-Bridge-page/.

Ingenious fabrication of bamboo leaf-like ferric vanadate intertwined multi-walled carbon nanotubes nanocomposite as a sensitive sensor for determination of uric acid in fetal bovine serum.

A novel nanocomposite material, ferric vanadate intertwined multi-walled carbon nanotubes (FeV/MWCNTs), has been designed which was drop-coated onto a glassy carbon electrode (GCE). The constructed sensor was used for the sensitive determination of uric acid (UA) in fetal bovine serum (FBS) and human serum (HS). A series of characterization and electrochemical tests showed that the ultrasound-assisted assembly of FeV with MWCNTs not only overcame the disadvantages of low conductivity and easy (unwanted) aggregation, but also avoided the decrease in effective surface area due to the severe aggregation of each individual raw material. The fabricated FeV/MWCNTs nanocomposites exhibited higher conductivity, larger effective surface area, and better electrocatalytic activity. In addition, under optimized conditions, the developed electrochemical sensor FeV/MWCNTs/GCE has a lower limit of detection (LOD, 0.05 µM; Ep = 0.268 V vs. Ag/AgCl) and wider linear range (0.20-100 µM), which can satisfy the criteria of trace UA detection. The results of UA determination in FBS (recovery = 95.5-103%; RSD ≤ 3.1%) and HS (recovery = 95.5-103%; RSD ≤ 4.3%) further validated the feasibility of FeV/MWCNTs-based electrochemical sensors for the determination of UA in biological fluids.

Efficient and Selective Removal of Organic Cationic Dyes by Peel of Brassica juncea Coss. var. gemmifera Lee et Lin-Based Biochar.

The design and preparation of cheaper, greener and more efficient adsorbents is essential for the removal of pollutants by adsorption. In this study, biochar was prepared from peel of Brassica juncea var. gemmifera Lee et Lin (PoBJ) using a facile, low-temperature and vacuum pyrolysis, and the adsorption mechanism toward organic dyes in aqueous solution was elucidated. The adsorbent was characterized by XPS, FT-IR and SEM, and zeta potential techniques. The adsorption ability of PoBJ biochar for cationic dyes (methylene blue, brilliant green, calcein-safranine, azure I, rhodamine B), anionic dyes (alizarin yellow R), and neutral dyes (neutral red) revealed that the biochar exhibited adsorption selectivity toward cationic dyes. The effects of different factors on the adsorption performance of PoBJ biochar, as well as the adsorption kinetics and thermodynamics, were further investigated by using methylene blue as the model adsorbate. These factors included temperature, pH, contact time and dye concentration. The experimental results showed that BJ280 and BJ160 (prepared at 280 °C and 160 °C, respectively) possessed relatively higher adsorption capacity of 192.8 and 167.40 mg g-1 for methylene blue (MB), respectively, demonstrating the possibility of utilization of PoBJ biochar as a superior bio-adsorbent. The experimental data of BJ160 toward MB were correlated with various kinetic and isothermal models. The results indicated that the adsorption process was consistent with the Langmuir isotherm model and nonlinear pseudo-second-order kinetic model. Thermodynamic parameters indicated that the adsorption of MB onto BJ160 was exothermic. Thus, the low-temperature prepared PoBJ biochar was an environmentally friendly, economic and efficient cationic dye adsorbent.

Bayesian Collaborative Learning for Whole-Slide Image Classification.

Whole-slide image (WSI) classification is fundamental to computational pathology, which is challenging in extra-high resolution, expensive manual annotation, data heterogeneity, etc. Multiple instance learning (MIL) provides a promising way towards WSI classification, which nevertheless suffers from the memory bottleneck issue inherently, due to the gigapixel high resolution. To avoid this issue, the overwhelming majority of existing approaches have to decouple the feature encoder and the MIL aggregator in MIL networks, which may largely degrade the performance. Towards this end, this paper presents a Bayesian Collaborative Learning (BCL) framework to address the memory bottleneck issue with WSI classification. Our basic idea is to introduce an auxiliary patch classifier to interact with the target MIL classifier to be learned, so that the feature encoder and the MIL aggregator in the MIL classifier can be learned collaboratively while preventing the memory bottleneck issue. Such a collaborative learning procedure is formulated under a unified Bayesian probabilistic framework and a principled Expectation-Maximization algorithm is developed to infer the optimal model parameters iteratively. As an implementation of the E-step, an effective quality-aware pseudo labeling strategy is also suggested. The proposed BCL is extensively evaluated on three publicly available WSI datasets, i.e., CAMELYON16, TCGA-NSCLC and TCGA-RCC, achieving an AUC of 95.6%, 96.0% and 97.5% respectively, which consistently outperforms all the methods compared. Comprehensive analysis and discussion will also be presented for in-depth understanding of the method. To promote future work, our source code is released at: https://github.com/Zero-We/BCL.

Decoration of alkalization-intercalated Ti3C2 with ZIF-8@ZIF-67-derived N-doped carbon nanocage for detecting 4-nitrophenol.

The highly effective alk-Ti3C2/bimetallic Co, Zn embedded N-doped carbon (Co-Zn-NC) composite was fabricated by a convenient self-assembled method strategy and applied to  the reduction of 4-nitrophenol(4-NP). Co-Zn-NC nanocage was synthesized by using designed core-shell ZIF-8@ZIF-67 as sacrificial template. The Co-Zn-NC was prepared by pyrolysis of ZIF-8@ZIF-67 at 900 °C with high-specific surface area and hollow structure, which facilitates the dispersion of Co species and produces abundant Co-Nx active sites. In addition, the electrochemical property and specific surface area of Ti3C2 were improved by alkaline treatment. As a result, compared with alk-Ti3C2 and Co-Zn-NC, the alk-Ti3C2/Co-Zn-NC sensor showed higher activity and stability in detecting 4-NP. The alk-Ti3C2/Co-Zn-NC sensor has a wide determination range of 2-500 µM and a low detection limit of 0.23 µM for 4-NP. In addition, the newly developed alk-Ti3C2/Co-Zn-NC sensor displayed satisfactory reproducibility and  good stability in detecting 4-NP in aqueous samples.

Prototypical multiple instance learning for predicting lymph node metastasis of breast cancer from whole-slide pathological images.

Computerized identification of lymph node metastasis of breast cancer (BCLNM) from whole-slide pathological images (WSIs) can largely benefit therapy decision and prognosis analysis. Besides the general challenges of computational pathology, like extra-high resolution, very expensive fine-grained annotation, etc., two particular difficulties with this task lie in (1) modeling the significant inter-tumoral heterogeneity in BCLNM pathological images, and (2) identifying micro-metastases, i.e., metastasized tumors with tiny foci. Towards this end, this paper presents a novel weakly supervised method, termed as Prototypical Multiple Instance Learning (PMIL), to learn to predict BCLNM from WSIs with slide-level class labels only. PMIL introduces the well-established vocabulary-based multiple instance learning (MIL) paradigm into computational pathology, which is characterized by utilizing the so-called prototypes to model pathological data and construct WSI features. PMIL mainly consists of two innovatively designed modules, i.e., the prototype discovery module which acquires prototypes from training data by unsupervised clustering, and the prototype-based slide embedding module which builds WSI features by matching constitutive patches against the prototypes. Relative to existing MIL methods for WSI classification, PMIL has two substantial merits: (1) being more explicit and interpretable in modeling the inter-tumoral heterogeneity in BCLNM pathological images, and (2) being more effective in identifying micro-metastases. Evaluation is conducted on two datasets, i.e., the public Camelyon16 dataset and the Zbraln dataset created by ourselves. PMIL achieves an AUC of 88.2% on Camelyon16 and 98.4% on Zbraln (at 40x magnification factor), which consistently outperforms other compared methods. Comprehensive analysis will also be carried out to further reveal the effectiveness and merits of the proposed method.

Removal of Chromium Species by Adsorption: Fundamental Principles, Newly Developed Adsorbents and Future Perspectives.

Emerging chromium (Cr) species have attracted increasing concern. A majority of Cr species, especially hexavalent chromium (Cr(VI)), could lead to lethal effects on human beings, animals, and aquatic lives even at low concentrations. One of the conventional water-treatment methodologies, adsorption, could remove these toxic Cr species efficiently. Additionally, adsorption possesses many advantages, such as being cost-saving, easy to implement, highly efficient and facile to design. Previous research has shown that the application of different adsorbents, such as carbon nanotubes (carbon nanotubes (CNTs) and graphene oxide (GO) and its derivatives), activated carbons (ACs), biochars (BCs), metal-based composites, polymers and others, is being used for Cr species removal from contaminated water and wastewater. The research progress and application of adsorption for Cr removal in recent years are reviewed, the mechanisms of adsorption are also discussed and the development trend of Cr treatment by adsorption is proposed.

Pose-Guided Hierarchical Semantic Decomposition and Composition for Human Parsing.

Human parsing is a fine-grained semantic segmentation task, which needs to understand human semantic parts. Most existing methods model human parsing as a general semantic segmentation, which ignores the inherent relationship among hierarchical human parts. In this work, we propose a pose-guided hierarchical semantic decomposition and composition framework for human parsing. Specifically, our method includes a semantic maintained decomposition and composition (SMDC) module and a pose distillation (PC) module. SMDC progressively disassembles the human body to focus on the more concise regions of interest in the decomposition stage and then gradually assembles human parts under the guidance of pose information in the composition stage. Notably, SMDC maintains the atomic semantic labels during both stages to avoid the error propagation issue of the hierarchical structure. To further take advantage of the relationship of human parts, we introduce pose information as explicit guidance for the composition. However, the discrete structure prediction in pose estimation is against the requirement of the continuous region in human parsing. To this end, we design a PC module to broadcast the maximum responses of pose estimation to form the continuous structure in the way of knowledge distillation. The experimental results on the look-into-person (LIP) and PASCAL-Person-Part datasets demonstrate the superiority of our method compared with the state-of-the-art methods, that is, 55.21% mean Intersection of Union (mIoU) on LIP and 69.88% mIoU on PASCAL-Person-Part.

A deep learning approach to the diagnosis of atelectasis and attic retraction pocket in otitis media with effusion using otoscopic images.

BACKGROUND: This study aimed to develop and validate a deep learning (DL) model to identify atelectasis and attic retraction pocket in cases of otitis media with effusion (OME) using multi-center otoscopic images. METHOD: A total of 6393 OME otoscopic images from three centers were used to develop and validate a DL model for detecting atelectasis and attic retraction pocket. A threefold random cross-validation procedure was adopted to divide the dataset into training validation sets on a patient level. A team of otologists was assigned to diagnose and characterize atelectasis and attic retraction pocket in otoscopic images. Receiver operating characteristic (ROC) curves, including area under the ROC curve (AUC), accuracy, sensitivity, and specificity were used to assess the performance of the DL model. Class Activation Mapping (CAM) illustrated the discriminative regions in the otoscopic images. RESULTS: Among all OME otoscopic images, 3564 (55.74%) were identified with attic retraction pocket, and 2460 (38.48%) with atelectasis. The diagnostic DL model of attic retraction pocket and atelectasis achieved a threefold cross-validation accuracy of 89% and 79%, AUC of 0.89 and 0.87, a sensitivity of 0.93 and 0.71, and a specificity of 0.62 and 0.84, respectively. Larger and deeper cases of atelectasis and attic retraction pocket showed greater weight, based on the red color depicted in the heat map of CAM. CONCLUSION: The DL algorithm could be employed to identify atelectasis and attic retraction pocket in otoscopic images of OME, and as a tool to assist in the accurate diagnosis of OME.

Three-dimensional visual technique based on CT lymphography data combined with methylene blue in endoscopic sentinel lymph node biopsy for breast cancer.

BACKGROUND: The combined application of blue dye and radioisotopes is currently the primary mapping technique used for sentinel lymph node biopsy (SLNB) in breast cancer patients. However, radiocolloid techniques have not been widely adopted, especially in developing countries, given the strict restrictions on radioactive materials. Consequently, we carried out a retrospective study to evaluate the feasibility and accuracy of three-dimensional visualization technique (3DVT) based on computed tomography-lymphography (CT-LG) in endoscopic sentinel lymph node biopsy (ESLNB) for breast cancer. METHODS: From September 2018 to June 2020, 389 patients who underwent surgical treatment of breast cancer in our department were included in this study. The CT-LG data of these patients were reconstructed into digital 3D models and imported into Smart Vision Works V1.0 to locate the sentinel lymph node (SLN) and for visual simulation surgery. ESLNB and endoscopic axillary lymph node dissection were carried out based on this new technique; the accuracy and clinical value of 3DVT in ESLNB were analyzed. RESULTS: The reconstructed 3D models clearly displayed all the structures of breast and axilla, which favors the intraoperative detection of SLNs. The identification rate of biopsied SLNs was 100% (389/389). The accuracy, sensitivity, and false-negative rate were 93.83% (365/389), 93.43% (128/137), and 6.57% (9/137), respectively. Upper limb lymphedema occurred in one patient 3 months after surgery during the 12-month follow-up period. CONCLUSIONS: Our 3DVT based on CT-LG data combined with methylene blue in ESLNB ensures a high identification rate of SLNs with low false-negative rates. It, therefore, has the potential to serve as a new method for SLN biopsy in breast cancer cases.

A Novel Low-Cost Bio-Sorbent Prepared from Crisp Persimmon Peel by Low-Temperature Pyrolysis for Adsorption of Organic Dyes.

In order to properly reuse food waste and remove various contaminants from wastewater, the development of green, sustainable and clean technologies has demonstrated potential in the efficient inhibition of secondary pollution to the environment. In this study, an economical and green method was used to prepare biochar from crisp persimmon peel (CPP) using flash-vacuum pyrolysis at different temperatures (200-700 °C; referred to as CPP200-CPP700). CPP200 has high polarity, low aromaticity and high oxygen-containing functional groups that exhibit superior MB adsorption capabilities. CPP200 that was prepared at a relatively low temperature of 200 °C exhibited a high adsorption capacity of 59.72 mg/g toward methylene blue (MB), which was relatively higher than that for alizarin yellow R (4.05 mg/g) and neutral red (39.08 mg/g), indicating that CPP200 possesses a higher adsorption selectivity for cationic dyes. Kinetics investigation revealed that the kinetic data of CPP200 for the adsorption of MB was better fitted by a linear pseudo-second-order model. Isothermal studies indicated that the linear Langmuir model was more suitable for describing the adsorption process. The adsorption thermodynamics illustrated that the adsorption of MB onto CPP200 was spontaneous and endothermic. EDS and IR analyses of CPP200 for both pre- and post-adsorption of MB showed that electrostatic interactions between oxygen-containing groups on biochar and target MB dominated the adsorption procedure, in addition to hydrogen bonding interactions. Reusability tests confirmed the excellent regeneration characteristics of CPP200, indicating that CPP200 may be used as a green, sustainable, highly efficient and recyclable adsorbent for the selective removal of cationic organic dyes.

Novel Silver-Plated Nickel-Coated Graphite Powder with Excellent Heat and Humidity Resistance: Facile Preparation and Performance Investigation.

Nickel-coated graphite (Ni/C) powder has many applications in diverse areas such as paint, print ink, adhesive, conductive rubber, and so on. To increase its stability in harsh environmental conditions, the electroless plating of silver film on Ni/C via ascorbic acid was studied. A silver layer with a thickness of 2.5 µm was successfully plated on Ni/C powder's surface with an Ag loading of 44.35 wt.%. Silica gel blended with the Ag/Ni/C powder exhibited much higher conductivity under aging conditions of 85 °C and 85% RH for 1000 h than that with pristine Ni/C powder. Further tests showed that the conductivity of Ag/Ni/C powder remained almost unchanged even in an extremely humid and hot condition for 1000 h. Aging tests were carried out for Ag/Ni/C and Ni/C powders under long-term humid and hot conditions (85 °C, 85% RH), in which Ag/Ni/C samples showed much better electromagnetic shielding performance. Due to the excellent properties and reasonable price, the potential applications of Ag/Ni/C in conductive glue and electromagnetic shielding glue could be expected.

Triallyl Isocyanurate as an Efficient Electrolyte Additive for Layered Oxide Cathode Material-Based Lithium-Ion Batteries with Improved Stability under High-Voltage.

In this study, a new electrolyte additive 1,3,5-tri-2-propenyl-1,3,5-triazine-2,4,6-(1H, 3H, 5H)-trione (TAIC) for lithium-ion batteries is reported. The additive is introduced as a novel electrolyte additive to enhance electrochemical performances of layered lithium nickel cobalt manganese oxide (NCM) and lithium cobalt oxide (LiCoO2) cathodes, especially under a higher working voltage. Encouragingly, we found protective films would be formed on the cathode surface by the electrochemical oxidation, and the stability of the cathode material-electrolyte interface was greatly promoted. By adding 0.5 wt.% of TAIC into the electrolyte, the battery exhibited outstanding performances. The thickness swelling decreased to about 6% after storage at 85 °C for 24 h, while the capacity retention of cycle-life performances under high temperature of 45 °C after the 600th cycle increased 10% in comparison with the batteries without TAIC. Due to its specific function, the additive can be used in high energy density and high voltage lithium-ion battery systems.

Novel lanthanum vanadate-based nanocomposite for simultaneously electrochemical detection of dopamine and uric acid in fetal bovine serum.

The abnormal levels of two biological molecules, dopamine (DA) and Uric acid (UA), in human body are symptoms of diseases such as Parkinson's disease and arrhythmia. A novel lanthanum vanadate and multi-walled carbon nanotubes (LaV-MWCNTs) composite modified glassy carbon electrode (GCE) was developed and utilized as an efficient electrochemical sensor for the simultaneous detection of DA and UA. LaV-MWCNTs composite was successfully fabricated by a facile ultrasonic self-assembly method and identified by means of a series of successive measurements including XPS, XRD, FT-IR and FE-SEM. The LaV-MWCNTs modified GCE shows the concentration linear ranges of DA and UA are 2-100 µΜ using DPV. The limits of detection (LODs; signal-to-noise ratio of 3, S/N = 3) of the LaV-MWCNTs modified GCE sensor for DA and UA were calculated to be 0.046 µM and 0.025 µM, respectively. The feasibility of using the LaV-MWCNTs modified GCE sensor to detect DA and UA in a typical biological fluid, fetal bovine serum, was also evaluated by the standard addition method.

A Critical Review on Black Phosphorus-Based Photocatalytic CO2 Reduction Application.

Energy shortages and greenhouse effects are two unavoidable problems that need to be solved. Photocatalytically converting CO2 into a series of valuable chemicals is considered to be an effective means of solving the above dilemmas. Among these photocatalysts, the utilization of black phosphorus for CO2 photocatalytic reduction deserves a lightspot not only for its excellent catalytic activity through different reaction routes, but also on account of the great preponderance of this relatively cheap catalyst. Herein, this review offers a summary of the recent advances in synthesis, structure, properties, and application for CO2 photocatalytic reduction. In detail, the review starts from the basic principle of CO2 photocatalytic reduction. In the following section, the synthesis, structure, and properties, as well as CO2 photocatalytic reduction process of black phosphorus-based photocatalyst are discussed. In addition, some possible influencing factors and reaction mechanism are also summarized. Finally, a summary and the possible future perspectives of black phosphorus-based photocatalyst for CO2 reduction are established.

Rational assembly of GO-based heterocyclic sulfur- and nitrogen-containing aerogels and their adsorption properties toward rare earth elementals.

An aromatic heterocyclic compound, 2-aminobenzothiazole (ABT), was used to decorate graphene oxide (GO) by a facile hydrothermal self-assembly procedure. The developed three-dimensional (3D) GO-ABT composite aerogels could be utilized as high-powered and sustainable adsorbents for the enrichment and recovery of low concentration rare earth elements (REEs) from aqueous solutions. The composition and microstructure of GO-ABT composites were explored various characterization methods. The enrichment properties of GO-ABT composites for REEs were investigated in detail, revealing the existence of S-, N- and -NH2 in ABT, as well as the carboxyl and hydroxyl groups of GO which might act as the major REE binding sites. The adsorption of GO-ABT composites for low concentration REEs could reach equilibrium in 30 min. Our investigations confirmed that the optimal pH value of GO-ABT composites for REEs was pH 4.0-5.0. For the adsorbent regeneration study, 50.0 mg of GO-ABT15:1/120 °C/6 h composite was used toward 20.0 mL of Er3+ solutions. After ten regeneration cycles, the adsorption rates of GO-ABT composites for Er3+ remained around 100%, and the desorption rates maintained over 90%. The long-term storage of the adsorbent did not affect its adsorption ability, while desorption rates increased, indicating it possessed relatively higher stability.

WS-RCNN: Learning to Score Proposals for Weakly Supervised Instance Segmentation.

Weakly supervised instance segmentation (WSIS) provides a promising way to address instance segmentation in the absence of sufficient labeled data for training. Previous attempts on WSIS usually follow a proposal-based paradigm, critical to which is the proposal scoring strategy. These works mostly rely on certain heuristic strategies for proposal scoring, which largely hampers the sustainable advances concerning WSIS. Towards this end, this paper introduces a novel framework for weakly supervised instance segmentation, called Weakly Supervised R-CNN (WS-RCNN). The basic idea is to deploy a deep network to learn to score proposals, under the special setting of weak supervision. To tackle the key issue of acquiring proposal-level pseudo labels for model training, we propose a so-called Attention-Guided Pseudo Labeling (AGPL) strategy, which leverages the local maximal (peaks) in image-level attention maps and the spatial relationship among peaks and proposals to infer pseudo labels. We also suggest a novel training loss, called Entropic OpenSet Loss, to handle background proposals more effectively so as to further improve the robustness. Comprehensive experiments on two standard benchmarking datasets demonstrate that the proposed WS-RCNN can outperform the state-of-the-art by a large margin, with an improvement of 11.6% on PASCAL VOC 2012 and 10.7% on MS COCO 2014 in terms of mAP50, which indicates that learning-based proposal scoring and the proposed WS-RCNN framework might be a promising way towards WSIS.

Learning Invariant Patterns Based on a Convolutional Neural Network and Big Electroencephalography Data for Subject-Independent P300 Brain-Computer Interfaces.

A brain-computer interface (BCI) measures and analyzes brain activity and converts this activity into computer commands to control external devices. In contrast to traditional BCIs that require a subject-specific calibration process before being operated, a subject-independent BCI learns a subject-independent model and eliminates subject-specific calibration for new users. However, building subject-independent BCIs remains difficult because electroencephalography (EEG) is highly noisy and varies by subject. In this study, we propose an invariant pattern learning method based on a convolutional neural network (CNN) and big EEG data for subject-independent P300 BCIs. The CNN was trained using EEG data from a large number of subjects, allowing it to extract subject-independent features and make predictions for new users. We collected EEG data from 200 subjects in a P300-based spelling task using two different types of amplifiers. The offline analysis showed that almost all subjects obtained significant cross-subject and cross-amplifier effects, with an average accuracy of more than 80%. Furthermore, more than half of the subjects achieved accuracies above 85%. These results indicated that our method was effective for building a subject-independent P300 BCI, with which more than 50% of users could achieve high accuracies without subject-specific calibration.

Improving the Stability of High-Voltage Lithium Cobalt Oxide with a Multifunctional Electrolyte Additive: Interfacial Analyses.

In recent years, various attempts have been made to meet the increasing demand for high energy density of lithium-ion batteries (LIBs). The increase in voltage can improve the capacity and the voltage platform performance of the electrode materials. However, as the charging voltage increases, the stabilization of the interface between the cathode material and the electrolyte will decrease, causing side reactions on both sides during the charge-discharge cycling, which seriously affects the high-temperature storage and the cycle performance of LIBs. In this study, a sulfate additive, dihydro-1,3,2-dioxathiolo[1,3,2]dioxathiole 2,2,5,5-tetraoxide (DDDT), was used as an efficient multifunctional electrolyte additive for high-voltage lithium cobalt oxide (LiCoO2). Nanoscale protective layers were formed on the surfaces of both the cathode and the anode electrodes by the electrochemical redox reactions, which greatly decreased the side reactions and improved the voltage stability of the electrodes. By adding 2% (wt.%) DDDT into the electrolyte, LiCoO2 exhibited improved Li-storage performance at the relatively high temperature of 60 °C, controlled swelling behavior (less than 10% for 7 days), and excellent cycling performance (capacity retention rate of 76.4% at elevated temperature even after 150 cycles).