Search

Multimodal reasoning is a critical component in the pursuit of artificial intelligence systems that exhibit human-like intelligence, especially when tackling complex tasks. While the chain-of-thought (CoT) technique has gained considerable attention, the existing ScienceQA dataset, which focuses on multimodal scientific questions and explanations from elementary and high school textbooks, lacks a comprehensive evaluation of diverse approaches. To address this gap, we present COCO Multi-Modal Reasoning Dataset(COCO-MMRD), a novel dataset that encompasses an extensive collection of open-ended questions, rationales, and answers derived from the large object dataset COCO. Unlike previous datasets that rely on multiple-choice questions, our dataset pioneers the use of open-ended questions in the context of multimodal CoT, introducing a more challenging problem that effectively assesses the reasoning capability of CoT models. Through comprehensive evaluations and detailed analyses, we provide valuable insights and propose innovative techniques, including multi-hop cross-modal attention and sentence-level contrastive learning, to enhance the image and text encoders. Extensive experiments demonstrate the efficacy of the proposed dataset and techniques, offering novel perspectives for advancing multimodal reasoning.

Optically addressable spin defects hosted in two-dimensional van der Waals materials represent a new frontier for quantum technologies, promising to lead to a new class of ultrathin quantum sensors and simulators. Recently, hexagonal boron nitride (hBN) has been shown to host several types of optically addressable spin defects, thus offering a unique opportunity to simultaneously address and utilise various spin species in a single material. Here we demonstrate an interplay between two separate spin species within a single hBN crystal, namely $S=1$ boron vacancy defects and visible emitter spins. We unambiguously prove that the visible emitters are $S=\frac{1}{2}$ spins and further demonstrate room temperature coherent control and optical readout of both spin species. Importantly, by tuning the two spin species into resonance with each other, we observe cross-relaxation indicating strong inter-species dipolar coupling. We then demonstrate magnetic imaging using the $S=\frac{1}{2}$ defects, both under ambient and cryogenic conditions, and leverage their lack of intrinsic quantization axis to determine the anisotropic magnetic susceptibility of a test sample. Our results establish hBN as a versatile platform for quantum technologies in a van der Waals host at room temperature.

Spatio-temporal predictive learning is a learning paradigm that enables models to learn spatial and temporal patterns by predicting future frames from given past frames in an unsupervised manner. Despite remarkable progress in recent years, a lack of systematic understanding persists due to the diverse settings, complex implementation, and difficult reproducibility. Without standardization, comparisons can be unfair and insights inconclusive. To address this dilemma, we propose OpenSTL, a comprehensive benchmark for spatio-temporal predictive learning that categorizes prevalent approaches into recurrent-based and recurrent-free models. OpenSTL provides a modular and extensible framework implementing various state-of-the-art methods. We conduct standard evaluations on datasets across various domains, including synthetic moving object trajectory, human motion, driving scenes, traffic flow and weather forecasting. Based on our observations, we provide a detailed analysis of how model architecture and dataset properties affect spatio-temporal predictive learning performance. Surprisingly, we find that recurrent-free models achieve a good balance between efficiency and performance than recurrent models. Thus, we further extend the common MetaFormers to boost recurrent-free spatial-temporal predictive learning. We open-source the code and models at https://github.com/chengtan9907/OpenSTL., 33 pages, 17 figures, 19 tables. Under review. For more details, please refer to https://github.com/chengtan9907/OpenSTL

Recent studies have shown competitive performance in protein design that aims to find the amino acid sequence folding into the desired structure. However, most of them disregard the importance of predictive confidence, fail to cover the vast protein space, and do not incorporate common protein knowledge. After witnessing the great success of pretrained models on diverse protein-related tasks and the fact that recovery is highly correlated with confidence, we wonder whether this knowledge can push the limits of protein design further. As a solution, we propose a knowledge-aware module that refines low-quality residues. We also introduce a memory-retrieval mechanism to save more than 50\% of the training time. We extensively evaluate our proposed method on the CATH, TS50, and TS500 datasets and our results show that our Knowledge-Design method outperforms the previous PiFold method by approximately 9\% on the CATH dataset. Specifically, Knowledge-Design is the first method that achieves 60+\% recovery on CATH, TS50 and TS500 benchmarks. We also provide additional analysis to demonstrate the effectiveness of our proposed method. The code will be publicly available.

PDF file - 1.7MB, DNA damage induced by UVR in the presence and absence of nutlin-3 as determined by gamma-H2AX immunofluorescence analyzed by flow cytometer

PDF file - 15MB, Hydrogen peroxide generation in melanocytes irradiated with 105 mJ/cm2 UVR in the presence and absence of pifithrin-alpha

PDF file - 281K, Dose-dependent accumulation of p53 in human melanocytes induced by 18 h incubation with increasing concentrations of nutlin-3

Epidermal melanocytes are skin cells specialized in melanin production. Activation of the melanocortin 1 receptor (MC1R) on melanocytes by α-melanocyte–stimulating hormone (α-MSH) induces synthesis of the brown/black pigment eumelanin that confers photoprotection from solar UV radiation (UVR). Contrary to keratinocytes, melanocytes are slow proliferating cells that persist in the skin for decades, in an environment with high levels of UVR-induced reactive oxygen species (ROS). We previously reported that in addition to its role in pigmentation, α-MSH also reduces oxidative stress and enhances the repair of DNA photoproducts in melanocytes, independent of melanin synthesis. Given the significance of ROS in carcinogenesis, here we investigated the mechanisms by which α-MSH exerts antioxidant effects in melanocytes. We show that activation of the MC1R by α-MSH contributes to phosphorylation of p53 on serine 15, a known requirement for stabilization and activation of p53, a major sensor of DNA damage. This effect is mediated by the cAMP/PKA pathway and by the activation of phosphoinositide 3-kinase (PI3K) ATR and DNA protein kinase (DNA-PK). α-MSH increases the levels of 8-oxoguanine DNA glycosylase (OGG1) and apurinic apyrimidinic endonuclease 1 (APE-1/Ref-1), enzymes essential for base excision repair. Nutlin-3, an HDM2 inhibitor, mimicked the effects of α-MSH resulting in reduced phosphorylation of H2AX (γ-H2AX), a marker of DNA damage. Conversely, the p53 inhibitor pifithrin-α or silencing of p53 abolished the effects of α-MSH and augmented oxidative stress. These results show that p53 is an important target of the downstream MC1R signaling that reduces oxidative stress and possibly malignant transformation of melanocytes. Mol Cancer Res; 10(6); 778–86. ©2012 AACR.

PDF file - 3.8MB, Western blott analysis - Absence of p63 isoforms in human melanocytes

PDF file - 364K, Down-regulation of p53 by shRNA and its effect on the expression of p21 in primary cultures of human melanocytes

Pretrained protein structure models without labels are crucial foundations for the majority of protein downstream applications. The conventional structure pretraining methods follow the mature natural language pretraining methods such as denoised reconstruction and masked language modeling but usually destroy the real representation of spatial structures. The other common pretraining methods might predict a fixed set of predetermined object categories, where a restricted supervised manner limits their generality and usability as additional labeled data is required to specify any other protein concepts. In this work, we introduce a novel unsupervised protein structure representation pretraining with a robust protein language model. In particular, we first propose to leverage an existing pretrained language model to guide structure model learning through an unsupervised contrastive alignment. In addition, a self-supervised structure constraint is proposed to further learn the intrinsic information about the structures. With only light training data, the pretrained structure model can obtain better generalization ability. To quantitatively evaluate the proposed structure models, we design a series of rational evaluation methods, including internal tasks (e.g., contact map prediction, distribution alignment quality) and external/downstream tasks (e.g., protein design). The extensive experimental results conducted on multiple tasks and specific datasets demonstrate the superiority of the proposed sequence-structure transformation framework.

Have you ever been troubled by the complexity and computational cost of SE(3) protein structure modeling and been amazed by the simplicity and power of language modeling? Recent work has shown promise in simplifying protein structures as sequences of protein angles; therefore, language models could be used for unconstrained protein backbone generation. Unfortunately, such simplification is unsuitable for the constrained protein inpainting problem, where the model needs to recover masked structures conditioned on unmasked ones, as it dramatically increases the computing cost of geometric constraints. To overcome this dilemma, we suggest inserting a hidden \textbf{a}tomic \textbf{d}irection \textbf{s}pace (\textbf{ADS}) upon the language model, converting invariant backbone angles into equivalent direction vectors and preserving the simplicity, called Seq2Direct encoder ($\text{Enc}_{s2d}$). Geometric constraints could be efficiently imposed on the newly introduced direction space. A Direct2Seq decoder ($\text{Dec}_{d2s}$) with mathematical guarantees is also introduced to develop a \textbf{SDS} ($\text{Enc}_{s2d}$+$\text{Dec}_{d2s}$) model. We apply the SDS model as the denoising neural network during the conditional diffusion process, resulting in a constrained generative model--\textbf{DiffSDS}. Extensive experiments show that the plug-and-play ADS could transform the language model into a strong structural model without loss of simplicity. More importantly, the proposed DiffSDS outperforms previous strong baselines by a large margin on the task of protein inpainting.

In this paper, we investigate the mean-square stabilization for discrete-time stochastic systems that endure both multiple input delays and multiplicative control-dependent noises. For such multi-delay stochastic systems, we for the first time put forward two stabilization criteria: Riccati type and Lyapunov type. On the one hand, we adopt a reduction method to reformulate the original multi-delay stochastic system to a delay-free auxiliary system and present their equivalent proposition for stabilization. Then, by introducing a delay-dependent algebraic Riccati equation (DDARE), we prove that the system under consideration is stabilizable if and only if the developed DDARE has a unique positive definite solution. On the other hand, we characterize the delay-dependent Lyapunov equation (DDLE)-based criterion, which can be verified by linear matrix inequality (LMI) feasibility test. Besides, under some restricted structure, we propose an existence theorem of delay margin and more importantly, derive an explicit formula for computing its exact value.

Additional file 2: Table S1. Primer name, sequence and annealing temperature used for the amplification of the three main avian haemosporidians included in our study, Plasmodium, Haemoproteus and Leucocytozoon. Primer length is included in the expected amplicon size.

While artificial intelligence has made remarkable strides in revealing the relationship between biological macromolecules' primary sequence and tertiary structure, designing RNA sequences based on specified tertiary structures remains challenging. Though existing approaches in protein design have thoroughly explored structure-to-sequence dependencies in proteins, RNA design still confronts difficulties due to structural complexity and data scarcity. Adding to the problem, direct transplantation of protein design methodologies into RNA design fails to achieve satisfactory outcomes although sharing similar structural components. In this study, we aim to systematically construct a data-driven RNA design pipeline. We crafted a large, well-curated benchmark dataset and designed a comprehensive structural modeling approach to represent the complex RNA tertiary structure. More importantly, we proposed a hierarchical data-efficient representation learning framework that learns structural representations through contrastive learning at both cluster-level and sample-level to fully leverage the limited data. By constraining data representations within a limited hyperspherical space, the intrinsic relationships between data points could be explicitly imposed. Moreover, we incorporated extracted secondary structures with base pairs as prior knowledge to facilitate the RNA design process. Extensive experiments demonstrate the effectiveness of our proposed method, providing a reliable baseline for future RNA design tasks. The source code and benchmark dataset will be released publicly.

Additional file 4: Figure S3. Uncropped electrophoresis gel of nested PCR of amplification of CytB gene of avian Plasmodium using Haemoproteus-specific primer set HaemF/AE982 producing amplicons of 820bp only. Cropped region presented in the manuscript is denoted by the red box and labelled as Fig. 2B.

A study was conducted to understand how psychoacoustics can play a role in music “listening”, particular in a small arts gallery. As the appreciation of a recorded audio of an exhibit can interfere with its adjacent exhibits, it is often very difficult to isolate individual exhibit’s audio display without extensive sound abatement or control measures especially in the low frequencies range. In our study, high-frequency audio is extracted and played through small speakers while its low-frequency component is channelled to a vibration shaker mounted beneath a standing false flooring located in front of the exhibit. By making use of haptics in vibrotactile and psychoacoustics, a person standing on the false flooring is still able to “hear” and perceive the broad frequency spectrum without losing much listening pleasure. The location of the shaker on the false flooring is optimised through vibration measurements for maximum vibrational amplitude response.

Additional file 1: Figure S1. Flowchart containing the workflow used for the detection of avian haemosporidians in our study.

Antibodies are crucial proteins produced by the immune system in response to foreign substances or antigens. The specificity of an antibody is determined by its complementarity-determining regions (CDRs), which are located in the variable domains of the antibody chains and form the antigen-binding site. Previous studies have utilized complex techniques to generate CDRs, but they suffer from inadequate geometric modeling. Moreover, the common iterative refinement strategies lead to an inefficient inference. In this paper, we propose a \textit{simple yet effective} model that can co-design 1D sequences and 3D structures of CDRs in a one-shot manner. To achieve this, we decouple the antibody CDR design problem into two stages: (i) geometric modeling of protein complex structures and (ii) sequence-structure co-learning. We develop a novel macromolecular structure invariant embedding, typically for protein complexes, that captures both intra- and inter-component interactions among the backbone atoms, including C$\alpha$, N, C, and O atoms, to achieve comprehensive geometric modeling. Then, we introduce a simple cross-gate MLP for sequence-structure co-learning, allowing sequence and structure representations to implicitly refine each other. This enables our model to design desired sequences and structures in a one-shot manner. Extensive experiments are conducted to evaluate our results at both the sequence and structure level, which demonstrate that our model achieves superior performance compared to the state-of-the-art antibody CDR design methods.

Additional file 5: Figure S4. Uncropped electrophoresis gel of nested PCR of amplification of CytB gene of avian Leucocytozoon using nested primer set HaemFL/HaemRL2 producing amplicon of 523 bp. Cropped regions presented in the manuscript is denoted by the red box and labelled as Fig. 2C.

A mixed passivation strategy is deployed to produce monodisperse and pure CsPbBr3 nanocrystals under ambient condition via post treatment using ZnBr2 in mixed organic solvents. This room temperature synthesis route provides precise size and shape control, and colloidal nanocubes or nanospheres can be obtained depending on the choice of organic solvents. Any secondary phases' nucleation was inhibited. The ZnBr2 can reduce the Br- vacancies at the surface termination layers, thus enhancing their properties. This approach can potentially offer a cheap and viable route for perovskite nanocrystals in optical and electrical devices. Agency for Science, Technology and Research (A*STAR) Published version The authors would like to thank the MOE2019-T2-2-032 and Monetary Academic Resources for Research (Grant No. 001561- 00001) in Nanyang Technological University, A∗ STAR Career Development Fund (Grant No. C210112054), and Singapore Struc tural Metal Alloy Program (Grant No. A18b1B0061).

Is there a unified model for generating molecules considering different conditions, such as binding pockets and chemical properties? Although target-aware generative models have made significant advances in drug design, they do not consider chemistry conditions and cannot guarantee the desired chemical properties. Unfortunately, merging the target-aware and chemical-aware models into a unified model to meet customized requirements may lead to the problem of negative transfer. Inspired by the success of multi-task learning in the NLP area, we use prefix embeddings to provide a novel generative model that considers both the targeted pocket's circumstances and a variety of chemical properties. All conditional information is represented as learnable features, which the generative model subsequently employs as a contextual prompt. Experiments show that our model exhibits good controllability in both single and multi-conditional molecular generation. The controllability enables us to outperform previous structure-based drug design methods. More interestingly, we open up the attention mechanism and reveal coupling relationships between conditions, providing guidance for multi-conditional molecule generation.

Additional file 3: Figure S2. Uncropped electrophoresis gel of nested-multiplex PCR amplification of CytB gene of avian Plasmodium and Haemoproteus using primer set AE980/AE982 and AE983/AE985 producing amplicons of 580bp only. A positive amplification of Haemoproteus is indicated by the 346 bp amplicon in the positive control. Cropped region presented in the manuscript is denoted by the red box and labelled as Fig. 2A.

Is there a unified framework for graph-based retrosynthesis prediction? Through analysis of full-, semi-, and non-template retrosynthesis methods, we discovered that they strive to strike an optimal balance between combinability and consistency: \textit{Should atoms be combined as motifs to simplify the molecular editing process, or should motifs be broken down into atoms to reduce the vocabulary and improve predictive consistency?} Recent works have studied several specific cases, while none of them explores different combinability-consistency trade-offs. Therefore, we propose MotifRetro, a dynamic motif editing framework for retrosynthesis prediction that can explore the entire trade-off space and unify graph-based models. MotifRetro comprises two components: RetroBPE, which controls the combinability-consistency trade-off, and a motif editing model, where we introduce a novel LG-EGAT module to dynamiclly add motifs to the molecule. We conduct extensive experiments on USPTO-50K to explore how the trade-off affects the model performance and finally achieve state-of-the-art performance.

Sign language recognition (SLR) is a weakly supervised task that annotates sign videos as textual glosses. Recent studies show that insufficient training caused by the lack of large-scale available sign datasets becomes the main bottleneck for SLR. Most SLR works thereby adopt pretrained visual modules and develop two mainstream solutions. The multi-stream architectures extend multi-cue visual features, yielding the current SOTA performances but requiring complex designs and might introduce potential noise. Alternatively, the advanced single-cue SLR frameworks using explicit cross-modal alignment between visual and textual modalities are simple and effective, potentially competitive with the multi-cue framework. In this work, we propose a novel contrastive visual-textual transformation for SLR, CVT-SLR, to fully explore the pretrained knowledge of both the visual and language modalities. Based on the single-cue cross-modal alignment framework, we propose a variational autoencoder (VAE) for pretrained contextual knowledge while introducing the complete pretrained language module. The VAE implicitly aligns visual and textual modalities while benefiting from pretrained contextual knowledge as the traditional contextual module. Meanwhile, a contrastive cross-modal alignment algorithm is designed to explicitly enhance the consistency constraints. Extensive experiments on public datasets (PHOENIX-2014 and PHOENIX-2014T) demonstrate that our proposed CVT-SLR consistently outperforms existing single-cue methods and even outperforms SOTA multi-cue methods., Comment: Accepted to CVPR 2023 (Highlight paper, 2.5% acceptance rate); Open source

The secondary structure of ribonucleic acid (RNA) is more stable and accessible in the cell than its tertiary structure, making it essential for functional prediction. Although deep learning has shown promising results in this field, current methods suffer from poor generalization and high complexity. In this work, we present RFold, a simple yet effective RNA secondary structure prediction in an end-to-end manner. RFold introduces a decoupled optimization process that decomposes the vanilla constraint satisfaction problem into row-wise and column-wise optimization, simplifying the solving process while guaranteeing the validity of the output. Moreover, RFold adopts attention maps as informative representations instead of designing hand-crafted features. Extensive experiments demonstrate that RFold achieves competitive performance and about eight times faster inference efficiency than the state-of-the-art method. The code and Colab demo are available in \href{http://github.com/A4Bio/RFold}{http://github.com/A4Bio/RFold}.

Particle swarm optimization (PSO) is a population-based random-search optimization technique that has become an increasingly important branch of swarm intelligence studies. Population diversity is an effective measurement that shows the distribution of particles in a search space. In this paper, we propose a diversity-based PSO algorithm that is combined with multiple techniques, and population diversity is used as a search-strategy selection criterion for particles. When diversity is high, we suggest implementing a search strategy that has a strong exploration ability. When diversity is low, we suggest implementing a search strategy that has a strong exploitation ability. In addition to our diversity-based method, we introduce a gradient-based local-search technique, multi-crossover operation, and disturbance strategy to help improve the performance of the proposed algorithm. In the experiments, we compare the proposed algorithm with 10 advanced PSO variants based on 40 widely used benchmark functions , including the CEC2017 benchmark. The results indicate that the proposed algorithm yields a better solution accuracy and convergence speed than those of other PSO variants.

The prediction of protein structures from sequences is an important task for function prediction, drug design, and related biological processes understanding. Recent advances have proved the power of language models (LMs) in processing the protein sequence databases, which inherit the advantages of attention networks and capture useful information in learning representations for proteins. The past two years have witnessed remarkable success in tertiary protein structure prediction (PSP), including evolution-based and single-sequence-based PSP. It seems that instead of using energy-based models and sampling procedures, protein language model (pLM)-based pipelines have emerged as mainstream paradigms in PSP. Despite the fruitful progress, the PSP community needs a systematic and up-to-date survey to help bridge the gap between LMs in the natural language processing (NLP) and PSP domains and introduce their methodologies, advancements and practical applications. To this end, in this paper, we first introduce the similarities between protein and human languages that allow LMs extended to pLMs, and applied to protein databases. Then, we systematically review recent advances in LMs and pLMs from the perspectives of network architectures, pre-training strategies, applications, and commonly-used protein databases. Next, different types of methods for PSP are discussed, particularly how the pLM-based architectures function in the process of protein folding. Finally, we identify challenges faced by the PSP community and foresee promising research directions along with the advances of pLMs. This survey aims to be a hands-on guide for researchers to understand PSP methods, develop pLMs and tackle challenging problems in this field for practical purposes.

Social media refers to any digital tool that allows users to quickly create and share content with the public. The COVID-19 pandemic has indirectly changed consumers' purchasing habits, causing them to shift from traditional store purchases to online retail store purchases. Social media also changes the specialized strategies among sellers and purchasers. The main objective of this study is to investigate the factors affecting consumer purchase behaviour through social media during the COVID-19 outbreak. A set of the self-administered questionnaire has been distributed to 215 targeted young adults in three states of Malaysia which are Johor, Melaka and Selangor. The four independent variables are price, convenience, product variety and risk, are tested on their relationship towards the dependent variable, which is the purchase behaviour of consumers during the COVID-19 outbreak. The collected data were keyed into SPSS version 26 and followed by using Partial Least Square Structural Equation Modeling (PLS-SEM 3.3.3) to assess the hypothesis. The analysis result showed that all hypotheses are supported. Lastly, the result of this research will benefit the marketers for their information to understand the consumer purchase behaviour through social media during the COVID-19 pandemic.

The digital and physical methods of establishing minimal curved surfaces are the basis for realizing the design of the minimal curved surface modeling structure. Based on this research background, the paper showed an artistic surface modeling method based on nonlinear over-limit difference equations. The article combines parameter optimization and 3D modeling methods to model the constructed surface modeling. The research found that the nonlinear out-of-limit difference equation proposed in the paper is more accurate than the standard fractional differential equation algorithm. For this reason, the method can be extended and applied to the design of artistic surface modeling.

The degradation of shear mechanical properties of rock fracture surfaces was determined after applying multiple dry-wet cycles. Artificially fractured feldspathic sandstone specimens were soaked in chemical solutions with pH values of 2, 7, and 12 for 3, 6, 9, 12, and 15 dry-wet cycles, followed by direct shear tests under normal stresses of 3, 6, 9, 12, and 15 MPa. The results showed that the pre-peak shear stiffness and peak shear strength of the fracture surfaces decreased, and the peak shear displacement increased progressively after cumulative dry-wet cycling treatments compared to the behavior of oven-dry rock fractures. Additionally, the pre-peak shear stiffness, peak shear strength, peak shear displacement, and residual shear strength decreased cumulatively as the number of dry-wet cycles increased. However, the chemistry of the wetting solution had little effect on mechanical behavior. Based on the Barton formula for describing the peak shear strength for rock fractures, an empirical formula for peak shear strength for irregular rock fractures under dry-wet cycling conditions is proposed by introducing a proportionality factor to describe the degree of deterioration of the rock fracture surface shear strength. The modified formula has a good fitting accuracy for the test shear strength data of sandstone fractures under dry-wet cycling conditions, which may assist in the practical estimation of the peak shear strength of rock fractures under dry-wet cycling conditions in engineering practice.

Can we inject the pocket-ligand interaction knowledge into the pre-trained model and jointly learn their chemical space? Pretraining molecules and proteins has attracted considerable attention in recent years, while most of these approaches focus on learning one of the chemical spaces and lack the injection of biological knowledge. We propose a co-supervised pretraining (CoSP) framework to simultaneously learn 3D pocket and ligand representations. We use a gated geometric message passing layer to model both 3D pockets and ligands, where each node's chemical features, geometric position and orientation are considered. To learn biological meaningful embeddings, we inject the pocket-ligand interaction knowledge into the pretraining model via contrastive loss. Considering the specificity of molecules, we further propose a chemical similarity-enhanced negative sampling strategy to improve the contrastive learning performance. Through extensive experiments, we conclude that CoSP can achieve competitive results in pocket matching, molecule property predictions, and virtual screening.

In this study, for the first time, a forward-rotating roller is proposed for the spreading of CF/PA12 composite powder in the selective laser sintering (SLS) process. The mesoscopic kinetic mechanism of composite particle spreading is investigated by utilizing the “multi-spherical” element within the discrete element method (DEM). The commercial software EDEM and the open-source DEM particle simulation code LIGGGHTS-PUBLIC are used for the simulations in this work. It is found that the forward-rotating roller produces a strong compaction on the powder pile than does the conventional counter-rotating roller, thus increasing the coordination number and mass flow rate of the particle flow, which significantly improves the powder bed quality. In addition, the forward-rotating pattern generates a braking friction force on the particles in the opposite direction to their spread, which affects the particle dynamics and deposition process. Therefore, appropriately increasing the roller rotation speed to make this force comparable to the roller dragging force could result in faster deposition of the composite particles to form a stable powder bed. This mechanism allows the forward-rotating roller to maintain a good powder bed quality, even at a high spreading speed, thus providing greater potential for the industry to improve the spreading efficiency of the SLS process.

From CNN, RNN, to ViT, we have witnessed remarkable advancements in video prediction, incorporating auxiliary inputs, elaborate neural architectures, and sophisticated training strategies. We admire these progresses but are confused about the necessity: is there a simple method that can perform comparably well? This paper proposes SimVP, a simple video prediction model that is completely built upon CNN and trained by MSE loss in an end-to-end fashion. Without introducing any additional tricks and complicated strategies, we can achieve state-of-the-art performance on five benchmark datasets. Through extended experiments, we demonstrate that SimVP has strong generalization and extensibility on real-world datasets. The significant reduction of training cost makes it easier to scale to complex scenarios. We believe SimVP can serve as a solid baseline to stimulate the further development of video prediction. The code is available at \href{https://github.com/gaozhangyang/SimVP-Simpler-yet-Better-Video-Prediction}{Github}.

Introduction: Cervical cancer is common cancer and ranked in fourth place in both incidence and mortality worldwide. It is 3rd most common female cancer in Malaysia with a lifetime risk of 1 in 116. Infection with high-risk oncogenic human papillomavirus (HPV) is recognized as one of the substantial risk factors for the development of cervical cancers. Methods: It was a cross-sectional study conducted to determine the prevalence of HPV infection and its subtypes among women with various degrees of abnormal smears, who were seen in the colposcopy clinic of Sarawak General Hospital within six months’ period from January to June 2018. We recruited 56 participants. There were 23 each for an atypical squamous cell of undetermined significance (ASC-US) and low-grade squamous intraepithelial lesion (LSIL) and 10 high- grade squamous intraepithelial lesion (HSIL). DNA was extracted, and HPV genotypes were determined via polymerase chain reaction (PCR) using two primer pairs MY09/MY11 and GP5+/GP6+. Results: The age ranged from 23 to 56 years, with a mean age of 42.96 years. HPV was detected in 20 out of 56 (35.7%). There were 6 high-risk oncogenic HPVs (18, 51, 52, 56, 58, 68) detected in participants and the most prevalent subtypes were 18, 52, and 58 (20% each). Four low-risk HPVs detected were 6, 53, 70, and 84. There was a significant association between the severity of cervical lesions and HPV positivity (P < 0.004). HSIL had the highest positive predictive value to have HPV infection as 70% compared to 43.4% of LSIL and 9.3% of ASC-US. Conclusion: Distribution of HPV subtypes from women with abnormal smears from Sarawak indicated a high prevalence of HPV 18, 52, and 58. We also identified HPV 70, which has never been reported in West Malaysia. These findings could contribute valuable information for HPV vaccination strategies, particularly for Sarawakian women.

Background: Oxadiazole fragment is one of the most prevalent structures in biochemicals, especially in the research of new pesticides. It is necessary to develop new insecticides with a different mode of action for the treatment of insecticide resistance problems. And, it is worth exploring the new active insecticidal lead structures with oxadiazole fragments. Methods: We used a “splicing up” method introducing the trifluoromethyl-oxadiazole moiety to 3- (ethylsulfonyl)-pyridine structure, and replaced the 6-position on the pyridine ring by different substituted amines. Then, a series of novel 3-(ethylsulfonyl)-pyridines containing trifluoromethyloxadiazole moiety were designed and synthesized. All these title compounds were confirmed by 1H NMR, 13C NMR and ESI-MS. Results: The primary insecticidal activity results indicated that some of them (A1-A7, A10, A13- A14) exhibited good mortality against Mythimna separate at 500 mg/L (80-100%), and compounds A13 and A14 have moderate insecticidal activity against M. separate at 250 mg/L (50-55%). Discussion: The bioassay results showed that the designed compounds did not achieve excellent insecticidal activity by introducing the potential oxadiazole fragment. Therefore, it seems that the special physicochemical properties of the oxadiazole fragment should be considered in fragment splicing-based design. Conclusion: According to the bioassay studies, the results revealed that compounds A13 and A14 may provide useful information for further designing of efficient insecticides.

The contribution the housing and construction sector makes to the national economy is closely tied to new project launches in the country. At present, nations are suffering from a global economic slowdown and a slumping property market. Governments and private developers have even initiated many affordable housing projects, but interestingly these units are not as popular as expected. Many housing projects across the world are initiated without a clear understanding of the needs and behavioural tendencies of homebuyers. The ability to understand homebuyers' behaviour is limited due to the lack of sufficient research using a concise and all-encompassing framework that is supported by a strong theoretical foundation. A thorough review and comparison of three fields of psychology found environmental psychology to be suitable and Mehrabian and Russell's (1974) theory to be relevant, and ought to be complemented with the theory of planned behaviour (TPB). A framework is proposed, positing that favourable evaluations of the environment (residential environment) and social factors (attitude, subjective norms, perceived behavioural control) will elicit favourable emotional states (pleasure, arousal, dominance/place attachment) which in turn lead to favourable behavioural response (homeownership intention). On the other hand, should the evaluations of the environment and social factors be unfavourable, this would lead to unfavourable emotions and subsequently unfavourable behavioural response (leasing intention).

Additively manufactured biodegradable zinc (Zn) alloy scaffolds constitute an important branch in orthopedic implants because of their moderate degradation behavior and bone-mimicking mechanical properties. This work investigated the corrosion fatigue response of a zinc-magnesium (Zn-Mg) alloy gyroid scaffold fabricated via laser-powder-bed-fusion additive manufacturing at the first time. The high-cycle compression-compression fatigue testing of the printed Zn-Mg scaffold was conducted in simulated body fluid, showing its favorable fatigue strength, structural reliability, and anti-fatigue capability. The printed Zn-Mg scaffold obtained a 227% higher fatigue strength than that of the printed Zn scaffold but 17% lower strain accumulation at 10

In this paper, we address the task of semantics-guided image outpainting, which is to complete an image by generating semantically practical content. Different from most existing image outpainting works, we approach the above task by understanding and completing image semantics at the scene graph level. In particular, we propose a novel network of Scene Graph Transformer (SGT), which is designed to take node and edge features as inputs for modeling the associated structural information. To better understand and process graph-based inputs, our SGT uniquely performs feature attention at both node and edge levels. While the former views edges as relationship regularization, the latter observes the co-occurrence of nodes for guiding the attention process. We demonstrate that, given a partial input image with its layout and scene graph, our SGT can be applied for scene graph expansion and its conversion to a complete layout. Following state-of-the-art layout-to-image conversions works, the task of image outpainting can be completed with sufficient and practical semantics introduced. Extensive experiments are conducted on the datasets of MS-COCO and Visual Genome, which quantitatively and qualitatively confirm the effectiveness of our proposed SGT and outpainting frameworks., CVPR 2022

The linear sequence of amino acids determines protein structure and function. Protein design, known as the inverse of protein structure prediction, aims to obtain a novel protein sequence that will fold into the defined structure. Recent works on computational protein design have studied designing sequences for the desired backbone structure with local positional information and achieved competitive performance. However, similar local environments in different backbone structures may result in different amino acids, indicating that protein structure's global context matters. Thus, we propose the Global-Context Aware generative de novo protein design method (GCA), consisting of local and global modules. While local modules focus on relationships between neighbor amino acids, global modules explicitly capture non-local contexts. Experimental results demonstrate that the proposed GCA method outperforms state-of-the-arts on de novo protein design. Our code and pretrained model will be released., ICASSP 2023

Over the last couple of decades, technological advancements have accelerated exponentially, especially in the realm of online social networking networks. The artificial intelligence (AI)-powered digital technologies applications continue to emerge to enhance and improve novel ways of communication on social media platforms, particularly Instagram. Indeed, this has caused a change in the behavioral and social customer journey, where customers need to embrace a digital experience adoption. The AI applications primarily aim to study the shoppers browsing trend to draw new clients and expand businesses. Even the fashion industry has tapped into Instagram's business benefits in this fast-paced and competitive industry. With this quick and compelling way to capture shoppers’ attention towards fashion products, the purchase decision may differ between e-shoppers and conventional shoppers. AI seems to be extremely promising and has the potential to be a game changer for Instagram users, advertisers, and influencers. This study applies the Engel-Kollat-Blackwell (EKB) theory to investigate the effects of AI-based digital technology experiences on Instagrammers’ fashion apparel purchase decisions - perceived eWOM, perceived emotional value, perceived quality, perceived risk and perceived price. Based on data collected from Instagram users, the framework of this study was evaluated using structural equation modelling (SEM). Semi-structured in-depth interviews were also conducted as part of the research to get a more in-depth understanding of the profiles and behaviors of Instagram users. Our findings from both methodologies confirm that perceived emotional value, perceived quality, and perceived eWOM revealed a statistically significant and positive influence on Instagrammers’ purchase decisions for fashion apparel. Meanwhile, the importance performance matrix analysis (IPMA) identified perceived emotional value as the most important factor for Instagrammers, but the highest performance was perceived quality. This research has important implications for Malaysian online retailers and shoppers to adapt to the fast-changing digital transformation. Assuredly, this study makes a noteworthy contribution to attitudinal research on social media commerce within the fashion industry.

Mixup is an efficient data augmentation approach that improves the generalization of neural networks by smoothing the decision boundary with mixed data. Recently, dynamic mixup methods have improved previous static policies effectively (e.g., linear interpolation) by maximizing salient regions or maintaining the target in mixed samples. The discrepancy is that the generated mixed samples from dynamic policies are more instance discriminative than the static ones, e.g., the foreground objects are decoupled from the background. However, optimizing mixup policies with dynamic methods in input space is an expensive computation compared to static ones. Hence, we are trying to transfer the decoupling mechanism of dynamic methods from the data level to the objective function level and propose the general decoupled mixup (DM) loss. The primary effect is that DM can adaptively focus on discriminative features without losing the original smoothness of the mixup while avoiding heavy computational overhead. As a result, DM enables static mixup methods to achieve comparable or even exceed the performance of dynamic methods. This also leads to an interesting objective design problem for mixup training that we need to focus on both smoothing the decision boundaries and identifying discriminative features. Extensive experiments on supervised and semi-supervised learning benchmarks across seven classification datasets validate the effectiveness of DM by equipping it with various mixup methods., The preprint revision, 15 pages, 6 figures. The source code is available at https://github.com/Westlake-AI/openmixup

The FNAL accelerator complex is poised to reach MW neutrino beams on target for the exploration of the dark sector physics and rare physics program spaces. Future operations of the complex will include CW linac operations at beam intensities that have not been seen before \cite{PIP2,RCS_LOI}. The ambitious beam program relies on multi-turn H$^{-}$ injection into the FNAL Booster and then extracted into delivery rings or the Booster Neutrino Beam (BNB) 8 GeV HEP program. A new rapid-cycling synchrotron (RCS) will be required to reach the LBNF goal of 2.4 MW because of intense space-charge limitations. There are many accelerator engineering challenges that are already known and many that will be discovered. This proposal calls for an intermediate step that will both facilitate the operation of Booster in the PIP-II era and gain operational experience associated with high power injection rings. This step includes the design, construction and installation of a 0.8 GeV accumulator ring (upgradeable to 1+ GeV) to be located in the PIP-II Booster Transfer Line (BTL). The PIP-II accumulator ring (PAR) may be primarily designed around permanent magnets or use standard iron core magnet technology with an aperture selected to accommodate the desired high intensity protons at 0.8 GeV.