Ultralong Cycling and Safe Lithium-Sulfur Pouch Cells for Sustainable Energy Storage.

While layered metal oxides remain the dominant cathode materials for the state-of-the-art lithium-ion batteries, conversion-type cathodes such as sulfur present unique opportunities in developing cheaper, safer, and more energy-dense next-generation battery technologies. There has been remarkable progress in advancing the laboratory scale lithium-sulfur (Li-S) coin cells to a high level of performance. However, the relevant strategies cannot be readily translated to practical cell formats such as pouch cells and even battery pack. Here these key technical challenges are addressed by molecular engineering of the Li metal for hydrophobicization, fluorination and thus favorable anode chemistry. The introduced tris(2,4-di-tert-butylphenyl) phosphite (TBP) and tetrabutylammonium fluoride (TBA+F-) as well as cellulose membrane by rolling enables the formation of a functional thin layer that eliminates the vulnerability of Li metal towards the already demanding environment required (1.55% relative humidity) for cell production and gives rise to LiF-rich solid electrolyte interphase (SEI) to suppress dendrite growth. As a result, Li-S pouch cells assembled at a pilot production line survive 400 full charge/discharge cycles with an average Coulombic efficiency of 99.55% and impressive rate performance of 1.5 C. A cell-level energy density of 417 Wh kg-1 and power density of 2766 W kg-1 are also delivered via multilayer Li-S pouch cell. The Li-S battery pack can even power an unmanned aerial vehicle of 3 kg for a fairly long flight time. This work represents a big step forward acceleration in Li-S battery marketization for future energy storage featuring improved safety, sustainability, higher energy density as well as reduced cost.

A Coding Framework and Benchmark Towards Low-Bitrate Video Understanding.

Video compression is indispensable to most video analysis systems. Despite saving the transportation bandwidth, it also deteriorates downstream video understanding tasks, especially at low-bitrate settings. To systematically investigate this problem, we first thoroughly review the previous methods, revealing that three principles, i.e., task-decoupled, label-free, and data-emerged semantic prior, are critical to a machine-friendly coding framework but are not fully satisfied so far. In this paper, we propose a traditional-neural mixed coding framework that simultaneously fulfills all these principles, by taking advantage of both traditional codecs and neural networks (NNs). On one hand, the traditional codecs can efficiently encode the pixel signal of videos but may distort the semantic information. On the other hand, highly non-linear NNs are proficient in condensing video semantics into a compact representation. The framework is optimized by ensuring that a transportation-efficient semantic representation of the video is preserved w.r.t. the coding procedure, which is spontaneously learned from unlabeled data in a self-supervised manner. The videos collaboratively decoded from two streams (codec and NN) are of rich semantics, as well as visually photo-realistic, empirically boosting several mainstream downstream video analysis task performances without any post-adaptation procedure. Furthermore, by introducing the attention mechanism and adaptive modeling scheme, the video semantic modeling ability of our approach is further enhanced. Fianlly, we build a low-bitrate video understanding benchmark with three downstream tasks on eight datasets, demonstrating the notable superiority of our approach. All codes, data, and models will be open-sourced for facilitating future research.

A New Biocontrol Agent Bacillus velezensis SF334 against Rubber Tree Fungal Leaf Anthracnose and Its Genome Analysis of Versatile Plant Probiotic Traits.

Natural rubber is an important national strategic and industrial raw material. The leaf anthracnose of rubber trees caused by the Colletotrichum species is one of the important factors restricting the yields of natural rubber. In this study, we isolated and identified strain Bacillus velezensis SF334, which exhibited significant antagonistic activity against both C. australisinense and C. siamense, the dominant species of Colletotrichum causing rubber tree leaf anthracnose in the Hainan province of China, from a pool of 223 bacterial strains. The cell suspensions of SF334 had a significant prevention effect for the leaf anthracnose of rubber trees, with an efficacy of 79.67% against C. siamense and 71.8% against C. australisinense. We demonstrated that SF334 can lead to the lysis of C. australisinense and C. siamense mycelia by causing mycelial expansion, resulting in mycelial rupture and subsequent death. B. velezensis SF334 also harbors some plant probiotic traits, such as secreting siderophore, protease, cellulase, pectinase, and the auxin of indole-3-acetic acid (IAA), and it has broad-spectrum antifungal activity against some important plant pathogenic fungi. The genome combined with comparative genomic analyses indicated that SF334 possesses most genes of the central metabolic and gene clusters of secondary metabolites in B. velezensis strains. To our knowledge, this is the first time a Bacillus velezensis strain has been reported as a promising biocontrol agent against the leaf anthracnose of rubber trees caused by C. siamense and C. australisinense. The results suggest that B. velezensis could be a potential candidate agent for the leaf anthracnose of rubber trees.

Lateral lymph node dissection can increase overall survival and 5­year survival rate of rectal cancer patients: A meta­analysis.

Rectal cancer is one of the most malignant tumors, and postoperative recurrence and metastasis are the main reasons for treatment failure. Lymph node metastasis is the main metastatic pathway of rectal cancer. The present study aimed to investigate the role of lateral lymph node dissection (LLND) in patients with rectal cancer using a meta-analysis. Articles in Chinese and English related to the application of LLND in patients with rectal cancer were retrieved and eligible studies were selected for data analysis. Evaluation indicators included the 5-year survival rate, recurrence rate, urinary system function and operation time. The random-effects model was utilized for the analysis. A total of 10 studies that met the eligibility criteria were selected, comprising 2,272 patients, including 1,101 cases in the LLND group and 1,171 cases in the non-LLND group. No significant difference was found between the two groups in terms of local recurrence rate, 5-year disease-free survival (DFS) rate, and DFS rate at the follow-up. It is noteworthy that cases in the LLND group had no significantly longer overall survival, but had a higher 5-year survival rate. However, cases in the LLND group had a longer operation time and worse urinary dysfunction. The results remained consistent throughout separate analyses for different research quality sources. The present meta-analysis showed that LLND provided a specific advantage in prolonging survival time. However, it was associated with prolonged operation time and an increased incidence of urinary dysfunction.

Nonflammable Polyfluorides-Anchored Quasi-Solid Electrolytes for Ultra-Safe Anode-Free Lithium Pouch Cells without Thermal Runaway.

The safe operation of rechargeable batteries is crucial because of numerous instances of fire and explosion mishaps. However, battery chemistry involving metallic lithium (Li) as the anode is prone to thermal runaway in flammable organic electrolytes under abusive conditions. Herein, an in situ encapsulation strategy is proposed to construct nonflammable quasi-solid electrolytes through the radical polymerization of a hexafluorobutyl acrylate (HFBA) monomer and a pentaerythritol tetraacrylate (PETEA) crosslinker. The quasi-solid system eliminates the inherent flammability of ether electrolytes with zero self-extinguishing time owing to the gas-phase radical capturing ability of HFBA. Additionally, the graphitized carbon layer generated during the decomposition of PETEA at high temperatures obstructs the heat and oxygen required for combustion. When coupled with Au-modified reduced graphene oxide anodic current collectors and lithium sulfide cathodes, the assembled anode-free Li-metal cell based on the quasi-solid electrolyte exhibits no signs of cell expansion or gas generation during cycling, and thermal runaway is eliminated under multiple mechanical, electrical, and thermal abuse scenarios and even rigorous strikes. This nonflammable quasi-solid configuration with gas- and condensed-phase flame-retardant mechanisms can drive a technological leap in anode-free Li-metal pouch cells and secure the practical applications necessary to power this society in a safe manner.

Antagonistic transcriptome profile reveals potential mechanisms of action on Xanthomonas oryzae pv. oryzicola by the cell-free supernatants of Bacillus velezensis 504, a versatile plant probiotic bacterium.

Bacterial leaf streak (BLS) of rice is a severe disease caused by the bacterial pathogen Xanthomonas oryzae pv. oryzicola (Xoc) that has gradually become the fourth major disease on rice in some rice-growing regions in southern China. Previously, we isolated a Bacillus velezensis strain 504 that exhibited apparent antagonistic activity against the Xoc wild-type strain RS105, and found that B. velezensis 504 was a potential biocontrol agent for BLS. However, the underlying mechanisms of antagonism and biocontrol are not completely understood. Here we mine the genomic data of B. velezensis 504, and the comparative transcriptomic data of Xoc RS105 treated by the cell-free supernatants (CFSs) of B. velezensis 504 to define differentially expressed genes (DEGs). We show that B. velezensis 504 shares over 89% conserved genes with FZB42 and SQR9, two representative model strains of B. velezensis, but 504 is more closely related to FZB42 than SQR9, as well as B. velezensis 504 possesses the secondary metabolite gene clusters encoding the essential anti-Xoc agents difficidin and bacilysin. We conclude that approximately 77% of Xoc RS105 coding sequences are differentially expressed by the CFSs of B. velezensis 504, which significantly downregulates genes involved in signal transduction, oxidative phosphorylation, transmembrane transport, cell motility, cell division, DNA translation, and five physiological metabolisms, as well as depresses an additional set of virulence-associated genes encoding the type III secretion, type II secretion system, type VI secretion system, type IV pilus, lipopolysaccharides and exopolysaccharides. We also show that B. velezensis 504 is a potential biocontrol agent for bacterial blight of rice exhibiting relative control efficiencies over 70% on two susceptible cultivars, and can efficiently antagonize against some important plant pathogenic fungi including Colletotrichum siamense and C. australisinense that are thought to be the two dominant pathogenic species causing leaf anthracnose of rubber tree in Hainan province of China. B. velezensis 504 also harbors some characteristics of plant growth-promoting rhizobacterium such as secreting protease and siderophore, and stimulating plant growth. This study reveals the potential biocontrol mechanisms of B. velezensis against BLS, and also suggests that B. velezensis 504 is a versatile plant probiotic bacterium.

Achieving Stable Lithium Metal Anode at 50 mA cm-2 Current Density by LiCl Enriched SEI.

Lithium metal batteries are intensively studied due to the potential to bring up breakthroughs in high energy density devices. However, the inevitable growth of dendrites will cause the rapid failure of battery especially under high current density. Herein, the utilization of tetrachloroethylene (C2 Cl4 ) is reported as the electrolyte additive to induce the formation of the LiCl-rich solid electrolyte interphase (SEI). Because of the lower Li ion diffusion barrier of LiCl, such SEI layer can supply sufficient pathway for rapid Li ion transport, alleviate the concentration polarization at the interface and inhibit the growth of Li dendrites. Meanwhile, the C2 Cl4 can be continuously replenished during the cycle to ensure the stability of the SEI layer. With the aid of C2 Cl4 -based electrolyte, the Li metal electrodes can maintain stable for >300 h under high current density of 50 mA cm-2 with areal capacity of 5 mAh cm-2 , broadening the compatibility of lithium metal anode toward practical application scenarios.

StyleVR: Stylizing Character Animations with Normalizing Flows.

The significance of artistry in creating animated virtual characters is widely acknowledged, and motion style is a crucial element in this process. There has been a long-standing interest in stylizing character animations with style transfer methods. However, this kind of models can only deal with short-term motions and yield deterministic outputs. To address this issue, we propose a generative model based on normalizing flows for stylizing long and aperiodic animations in the VR scene. Our approach breaks down this task into two sub-problems: motion style transfer and stylized motion generation, both formulated as the instances of conditional normalizing flows with multi-class latent space. Specifically, we encode high-frequency style features into the latent space for varied results and control the generation process with style-content labels for disentangled edits of style and content. We have developed a prototype, StyleVR, in Unity, which allows casual users to apply our method in VR. Through qualitative and quantitative comparisons, we demonstrate that our system outperforms other methods in terms of style transfer as well as stochastic stylized motion generation.

CLSA: A Contrastive Learning Framework with Selective Aggregation for Video Rescaling.

Video rescaling has recently drawn extensive attention for its practical applications such as video compression. Compared to video super-resolution, which focuses on upscaling bicubic-downscaled videos, video rescaling methods jointly optimize a downscaler and a upscaler. However, the inevitable loss of information during downscaling makes the upscaling procedure still ill-posed. Furthermore, the network architecture of previous methods mostly relies on convolution to aggregate information within local regions, which cannot effectively capture the relationship between distant locations. To address the above two issues, we propose a unified video rescaling framework by introducing the following designs. First, we propose to regularize the information of the downscaled videos via a contrastive learning framework, where, particularly, hard negative samples for learning are synthesized online. With this auxiliary contrastive learning objective, the downscaler tends to retain more information that benefits the upscaler. Second, we present a selective global aggregation module (SGAM) to efficiently capture long-range redundancy in high-resolution videos, where only a few representative locations are adaptively selected to participate in the computationally-heavy self-attention (SA) operations. SGAM enjoys the efficiency of the sparse modeling scheme while preserving the global modeling capability of SA. We refer to the proposed framework as Contrastive Learning framework with Selective Aggregation (CLSA) for video rescaling. Comprehensive experimental results show that CLSA outperforms video rescaling and rescaling-based video compression methods on five datasets, achieving state-of-the-art performance.

A Physical Organic Chemistry Approach to Developing Cyclopropenium-Based Energy Storage Materials for Redox Flow Batteries.

ConspectusRedox flow batteries (RFBs) represent a promising modality for electrical energy storage. In these systems, energy is stored via paired redox reactions of molecules on opposite sides of an electrochemical cell. Thus, a central objective for the field is to design molecules with the optimal combination of properties to serve as energy storage materials in RFBs. The ideal molecules should undergo reversible redox reactions at relatively high potentials (for the molecule that is oxidized during battery charging, called the catholyte) or low potentials (for the species that is reduced during battery charging, called the anolyte). Furthermore, anolytes and catholytes must be highly soluble in the electrolyte solution and stable to extended electrochemical cycling in all battery-relevant redox states. The ideal candidates would undergo more than one reversible electron transfer event. Finally, the optimal structures should be resistant to crossover through a selective separator in order to maintain isolation of the two sides of the cell. This Account describes our design and optimization of organic molecules for this application. We first provide background for the metrics and experiments used to characterize anolytes/catholytes and to progress them toward deployment in flow batteries. We then use our studies of aminocyclopropenium-based catholytes to illustrate this workflow and approach.We identified tris(dimethylamino) cyclopropenium hexafluorophosphate as a first-generation catholyte for nonaqueous RFBs based on literature reports from the 1970s describing its reversible chemical and electrochemical oxidation. Cyclic voltammetry and electrochemical cycling experiments in acetonitrile/LiPF6 confirmed that this molecule undergoes oxidation at relatively high potential (0.86 V versus ferrocene/ferrocenium) and exhibits moderate stability toward charge-discharge cycling. Replacing the methyl groups with isopropyl substituents led to enhanced cycling stability but poor solubility of the radical dication (<0.1 M in acetonitrile). Solubility was optimized using quantitative structure-property relationship modeling, which predicted derivatives with ≥10-fold enhanced solubility. Cyclopropeniums with 300-500 mV higher redox potentials were identified by replacing one of the dialkylamino substituents with a less electron-donating thioalkyl or aryl group. Multielectron catholytes were developed by creating hybrid structures that contain a di(amino) cyclopropenium conjugated with a phenothiazine moeity. Finally, oligomeric tris(amino) cyclopropeniums were designed as crossover resistant catholytes. Optimization of their solubility enabled the deployment of these oligomers in high concentration asymmetric redox flow batteries with energy densities that are comparable to the state-of-the-art commercial aqueous inorganic systems.

The natural pyrazolotriazine pseudoiodinine from Pseudomonas mosselii 923 inhibits plant bacterial and fungal pathogens.

Natural products largely produced by Pseudomonads-like soil-dwelling microorganisms are a consistent source of antimicrobial metabolites and pesticides. Herein we report the isolation of Pseudomonas mosselii strain 923 from rice rhizosphere soils of paddy fields, which specifically inhibit the growth of plant bacterial pathogens Xanthomonas species and the fungal pathogen Magnaporthe oryzae. The antimicrobial compound is purified and identified as pseudoiodinine using high-resolution mass spectra, nuclear magnetic resonance and single-crystal X-ray diffraction. Genome-wide random mutagenesis, transcriptome analysis and biochemical assays define the pseudoiodinine biosynthetic cluster as psdABCDEFG. Pseudoiodinine biosynthesis is proposed to initiate from guanosine triphosphate and 1,6-didesmethyltoxoflavin is a biosynthetic intermediate. Transposon mutagenesis indicate that GacA is the global regulator. Furthermore, two noncoding small RNAs, rsmY and rsmZ, positively regulate pseudoiodinine transcription, and the carbon storage regulators CsrA2 and CsrA3, which negatively regulate the expression of psdA. A 22.4-fold increase in pseudoiodinine production is achieved by optimizing the media used for fermentation, overexpressing the biosynthetic operon, and removing the CsrA binding sites. Both of the strain 923 and purified pseudoiodinine in planta inhibit the pathogens without affecting the rice host, suggesting that pseudoiodinine can be used to control plant diseases.

Learning Multi-Attention Context Graph for Group-Based Re-Identification.

Learning to re-identify or retrieve a group of people across non-overlapped camera systems has important applications in video surveillance. However, most existing methods focus on (single) person re-identification (re-id), ignoring the fact that people often walk in groups in real scenarios. In this work, we take a step further and consider employing context information for identifying groups of people, i.e., group re-id. On the one hand, group re-id is more challenging than single person re-id, since it requires both a robust modeling of local individual person appearance (with different illumination conditions, pose/viewpoint variations, and occlusions), as well as full awareness of global group structures (with group layout and group member variations). On the other hand, we believe that person re-id can be greatly enhanced by incorporating additional visual context from neighboring group members, a task which we formulate as group-aware (single) person re-id. In this paper, we propose a novel unified framework based on graph neural networks to simultaneously address the above two group-based re-id tasks, i.e., group re-id and group-aware person re-id. Specifically, we construct a context graph with group members as its nodes to exploit dependencies among different people. A multi-level attention mechanism is developed to formulate both intra-group and inter-group context, with an additional self-attention module for robust graph-level representations by attentively aggregating node-level features. The proposed model can be directly generalized to tackle group-aware person re-id using node-level representations. Meanwhile, to facilitate the deployment of deep learning models on these tasks, we build a new group re-id dataset which contains more than 3.8K images with 1.5K annotated groups, an order of magnitude larger than existing group re-id datasets. Extensive experiments on the novel dataset as well as three existing datasets clearly demonstrate the effectiveness of the proposed framework for both group-based re-id tasks.

Single Nickel Atom Catalysts Enable Fast Polysulfide Redox for Safe and Long-Cycle Lithium-Sulfur Batteries.

Lithium-sulfur (Li-S) batteries have attracted great interest due to their low cost, high theoretical energy density, and environmental friendliness. However, the sluggish conversion of lithium polysulfides (LiPS) to S and Li2 S during the charge/discharge process leads to unsatisfactory rate performance of lower to 0.1 C (1 C = 1675 mA g-1 ) especially for Li-S pouch batteries, thus hindering their practical applications in high power batteries. Here, well-defined and monodispersed Ni single-atom catalysts (SACs) embedded in highly porous nitrogen-doped graphitic carbons (NiSA-N-PGC) are designed and synthesized to form Ni-N4 catalytic sites at the atomic level. When serving as a bifunctional electrocatalyst, the Ni-N4 catalytic sites cannot only promote the interfacial conversion redox of LiPS by accelerating the transformation kinetics, but also suppress the undesirable shuttle effect by immobilizing LiPS. These findings are verified by both experimental results and DFT theoretical calculations. Furthermore, Li ions show low diffusion barrier on the surface of Ni-N4 sites, resulting in enhanced areal capacity of batteries. As a result, the Li-S battery delivers stable cycling life of more than 600 cycles with 0.069% capacity decay per cycle at a rate of 0.5 C. More importantly, the Li-S pouch cells with NiSA-N-PGC show an initial capacity of 1299 mAh g-1 at a rate of 0.2 C even with high sulfur loading of 6 mg cm-2 . This work opens up an avenue for developing single-atom catalysts to accelerate the kinetic conversion of LiPS for highly stable Li-S batteries.

Tailoring Lithium Fluoride Interface for Dendrite-Free Lithium Anode to Prolong the Cyclic Stability of Lithium-Sulfur Pouch Cells.

Lithium-sulfur (Li-S) cells have been regarded as attractive alternatives to achieve higher energy densities because of their theoretical specific energy far beyond the lithium-ion cells. However, the achieved results of Li-S cells are exaggerating the cycle performance in their pouch formats because the considerable works are based on the coin cells where flood electrolyte and endless Li supply ensure the Li metal with nature structure features, resulting in a negligible effect on cycle performance caused by the Li dendrites and electrolyte dissipation during cycles. Herein, we demonstrate a strategy to enable the Li metal with lithium fluoride (LiF)-rich solid electrolyte interface via integrating a reinforced interface (RI) embedded with nano-LiF particles on the surface of the Li metal anode. The RI interface enables the solvent molecules of the electrolyte to gain fewer electrons from Li anode, resulting in a lower leakage current of assembled RI||Li-S cell (~ 0 µA) than pristine Li anode (~ 1.15 µA). Moreover, these results show that suppressing lithium dendrite growth is more urgent than inhibiting the shuttle effect of polysulfides in the pouch cell format. As a result, the RI layer-engineered Li metal bears witness to the cyclic stability of Li anode over 800 h, thus achieving stable cycles of Ah-scale Li-S pouch cell with an energy density of 410 Wh/kg at a current of 200 mA per cell. Our study demonstrates that the suppression of lithium dendrites by the RI could be a promising method to prolong the cycle number of Li-S pouch cells.

High Energy Density, Asymmetric, Nonaqueous Redox Flow Batteries without a Supporting Electrolyte.

Energy density in nonaqueous redox flow batteries (RFBs) is often limited by the modest solubility of the redox-active organic molecules (ROMs). In addition, the lack of a separator that prevents ROMs from crossing between anolyte and catholyte solutions necessitates the use of 1:1 mixtures of two ROMs in both the anolyte and catholyte solutions in symmetric RFBs, further limiting concentrations. We show that permanently cationic oligomers of viologen, tris(dialkylamino)cyclopropenium, and phenothiazine molecules have high solubility in acetonitrile and cross over an anion exchange membrane at slow to undetectable rates, enabling the creation of asymmetric RFBs with low crossover. No added supporting electrolyte is necessary, with only the PF6- counteranions of the ROMs crossing the membrane during charge/discharge. An oligomeric viologen + oligomeric cyclopropenium RFB at 1.0 M (redox equivalents) has a voltage of 1.66 V and a theoretical energy density of 22.2 Wh/L, one of the highest reported for nonaqueous RFBs.

The MinCDE Cell Division System Participates in the Regulation of Type III Secretion System (T3SS) Genes, Bacterial Virulence, and Motility in Xanthomonas oryzae pv. oryzae.

Xanthomonas oryzae pv. oryzae (Xoo) causes bacterial leaf blight (BLB) in rice, which is one of the most severe bacterial diseases in rice in some Asian countries. The type III secretion system (T3SS) of Xoo encoded by the hypersensitive response and pathogenicity (hrp) genes is essential for its pathogenicity in host rice. Here, we identified the Min system (MinC, MinD, and MinE), a negative regulatory system for bacterial cell division encoded by minC, minD, and minE genes, which is involved in negative regulation of hrp genes (hrpB1 and hrpF) in Xoo. We found that the deletion of minC, minD, and minCDE resulted in enhanced hrpB1 and hrpF expression, which is dependent on two key hrp regulators HrpG and HrpX. The minC, minD, and minCDE mutants exhibited elongated cell lengths, and the classic Min system-defective cell morphology including minicells and short filamentations. Mutation of minC in Xoo resulted in significantly impaired virulence in host rice, swimming motility, and enhanced biofilm formation. Our transcriptome profiling also indicated some virulence genes were differentially expressed in the minC mutants. To our knowledge, this is the first report about the Min system participating in the regulation of T3SS expression. It sheds light on the understanding of Xoo virulence mechanisms.

Mapping Techniques for the Design of Lithium-Sulfur Batteries.

Mapping technique has been the powerful tool for the design of next-generation energy storage devices. Unlike the traditional ion-insertion based lithium batteries, the Li-S battery is based on the complex conversion reactions, which require more cooperation from mapping techniques to elucidate the underlying mechanism. Therefore, in this review, the representative works of mapping techniques for Li-S batteries are summarized, and categorized into the studies of lithium metal anode and sulfur cathode, with sub-sections based on shared characterization mechanisms. Due to specific features of mapping techniques, various aspects such as compositional distribution, in-plain/cross section characterization, coin cell/pouch cell configuration, and structural/mechanical analysis are emphasized in each study, aiming for the guidance for developing strategies to improve the battery performances. Benefited from the achieved progresses, suggestions for future studies based on mapping techniques are proposed to accelerate the development and commercialization of the Li-S battery.

Fine-Grained Video Captioning via Graph-based Multi-Granularity Interaction Learning.

Learning to generate continuous linguistic descriptions for multi-subject interactive videos in great details has particular applications in team sports auto-narrative. In contrast to traditional video caption, this task is more challenging as it requires simultaneous modeling of fine-grained individual actions, uncovering of spatio-temporal dependency structures of frequent group interactions, and then accurate mapping of these complex interaction details into long and detailed commentary. To explicitly address these challenges, we propose a novel framework Graph-based Learning for Multi-Granularity Interaction Representation (GLMGIR) for fine-grained team sports auto-narrative task. A multi-granular interaction modeling module is proposed to extract among-subjects' interactive actions in a progressive way for encoding both intra- and inter-team interactions. Based on the above multi-granular representations, a multi-granular attention module is developed to consider action/event descriptions of multiple spatio-temporal resolutions. Both modules are integrated seamlessly and work in a collaborative way to generate the final narrative. In the meantime, to facilitate reproducible research, we collect a new video dataset from YouTube.com called Sports Video Narrative dataset (SVN). It is a novel direction as it contains 6K team sports videos (i.e., NBA basketball games) with 10K ground-truth narratives(e.g., sentences). Furthermore, as previous metrics such as METEOR (i.e., used in coarse-grained video caption task) DO NOT cope with fine-grained sports narrative task well, we hence develop a novel evaluation metric named Fine-grained Captioning Evaluation (FCE), which measures how accurate the generated linguistic description reflects fine-grained action details as well as the overall spatio-temporal interactional structure. Extensive experiments on our SVN dataset have demonstrated the effectiveness of the proposed framework for fine-grained team sports video auto-narrative.

Bactericidal Effect of Pseudomonas oryziphila sp. nov., a Novel Pseudomonas Species Against Xanthomonas oryzae Reduces Disease Severity of Bacterial Leaf Streak of Rice.

Pseudomonas is a diverse genus of Gammaproteobacteria with increasing novel species exhibiting versatile trains including antimicrobial and insecticidal activity, as well as plant growth-promoting, which make them well suited as biocontrol agents of some pathogens. Here we isolated strain 1257 that exhibited strong antagonistic activity against two pathovars of Xanthomonas oryzae, especially X. oryzae pv. oryzicola (Xoc) responsible for the bacterial leaf streak (BLS) in rice. The phylogenetic, genomic, physiological, and biochemical characteristics support that strain 1257 is a representative of a novel Pseudomonas species that is most closely related to the entomopathogenic bacterium Pseudomonas entomophila. We propose to name it Pseudomonas oryziphila sp. nov. Comparative genomics analyses showed that P. oryziphila 1257 possesses most of the central metabolic genes of two closely related strains P. entomophila L48 and Pseudomonas mosselii CFML 90-83, as well as a set of genes encoding the type IV pilus system, suggesting its versatile metabolism and motility properties. Some features, such as insecticidal toxins, phosphate solubilization, indole-3-acetic acid, and phenylacetic acid degradation, were disclosed. Genome-wide random mutagenesis revealed that the non-ribosomal peptide catalyzed by LgrD may be a major active compound of P. oryziphila 1257 against Xoc RS105, as well as the critical role of the carbamoyl phosphate and the pentose phosphate pathway that control the biosynthesis of this target compound. Our findings demonstrate that 1257 could effectively inhibit the growth and migration of Xoc in rice tissue to prevent the BLS disease. To our knowledge, this is the first report of a novel Pseudomonas species that displays a strong antibacterial activity against Xoc. The results suggest that the P. oryziphila strain could be a promising biological control agent for BLS.

Development of High Energy Density Diaminocyclopropenium-Phenothiazine Hybrid Catholytes for Non-Aqueous Redox Flow Batteries.

This report describes the design of diaminocyclopropenium-phenothiazine hybrid catholytes for non-aqueous redox flow batteries. The molecules are synthesized in a rapid and modular fashion by appending a diaminocyclopropenium (DAC) substituent to the nitrogen of the phenothiazine. Combining a versatile C-N coupling protocol (which provides access to diverse derivatives) with computation and structure-property analysis enabled the identification of a catholyte that displays stable two-electron cycling at potentials of 0.64 and 1.00 V vs. Fc/Fc+ as well as high solubility in all oxidation states (≥0.45 M in TBAPF6 /MeCN). This catholyte was deployed in a high energy density two-electron RFB, exhibiting >90 % capacity retention over 266 hours of flow cell cycling at >0.5 M electron concentration.