Explore the vast range of titles available.

Temporal action detection is a very important and challenging task in the field of video understanding, especially for datasets with significant differences in action duration. The temporal relationships between the action instances contained in these datasets are very complex. For such videos, it is necessary to capture information with a richer temporal distribution as much as possible. In this paper, we propose a dual-stream model that can model contextual information at multiple temporal scales. First, the input video is divided into two resolution streams, followed by a Multi-Resolution Context Aggregation module to capture multi-scale temporal information. Additionally, an Information Enhancement module is added after the high-resolution input stream to model both long-range and short-range contexts. Finally, the outputs of the two modules are merged to obtain features with rich temporal information for action localization and classification. We conducted experiments on three datasets to evaluate the proposed approach. On ActivityNet-v1.3, an average mAP (mean Average Precision) of 32.83% was obtained. On Charades, the best performance was obtained, with an average mAP of 27.3%. On TSU (Toyota Smarthome Untrimmed), an average mAP of 33.1% was achieved.

Background Medicago sativa is the most important forage world widely, and is characterized by high quality and large biomass. While abiotic factors such as salt stress can negatively impact the growth and productivity of alfalfa. Maintaining Na + /K + homeostasis in the cytoplasm helps reduce cell damage and nutritional deprivation, which increases a salt-tolerance of plant. Teosinte Branched1/ Cycloidea/ Proliferating cell factors (TCP) family genes, a group of plant-specific transcription factors (TFs), involved in regulating plant growth and development and abiotic stresses. Recent studies have shown TCPs control the Na + /K + concentration of plants during salt stress. In order to improve alfalfa salt tolerance, it is important to identify alfalfa TCP genes and investigate if and how they regulate alfalfa Na + /K + homeostasis. Results Seventy-one MsTCP s including 23 non-redundant TCP genes were identified in the database of alfalfa genome (C.V XinJiangDaYe), they were classified into class I PCF (37 members) and class II: CIN (28 members) and CYC/TB1 (9 members). Their distribution on chromosome were unequally. MsTCP s belonging to PCF were expressed specifically in different organs without regularity, which belonging to CIN class were mainly expressed in mature leaves. MsTCP s belongs to CYC/TB1 clade had the highest expression level at meristem. Cis-elements in the promoter of MsTCP s were also predicted, the results indicated that most of the MsTCP s will be induced by phytohormone and stress treatments, especially by ABA-related stimulus including salinity stress. We found 20 out of 23 MsTCP s were up-regulated in 200 mM NaCl treatment, and MsTCP3/14/15/18 were significantly induced by 10 μM KCl, a K + deficiency treatment. Fourteen non-redundant MsTCPs contained miR319 target site, 11 of them were upregulated in MIM319 transgenic alfalfa, and among them four ( MsTCP3/4/10A/B ) genes were directly degraded by miR319. MIM319 transgene alfalfa plants showed a salt sensitive phenotype, which caused by a lower content of potassium in alfalfa at least partly. The expression of potassium transported related genes showed significantly higher expression in MIM319 plants. Conclusions We systematically analyzes the MsTCP gene family at a genome-wide level and reported that miR319- TCPs model played a function in K + up-taking and/ or transportation especially in salt stress. The study provide valuable information for future study of  TCP  genes in alfalfa and supplies candidate genes for salt-tolerance alfalfa molecular-assisted breeding.

Artificial chiral materials and nanostructures with strong and tuneable chiroptical activities, including sign, magnitude, and wavelength distribution, are useful owing to their potential applications in chiral sensing, enantioselective catalysis, and chiroptical devices. Thus, the inverse design and customized manufacturing of these materials is highly desirable. Here, we use an artificial intelligence (AI) guided robotic chemist to accurately predict chiroptical activities from the experimental absorption spectra and structure/process parameters, and generate chiral films with targeted chiroptical activities across the full visible spectrum. The robotic AI-chemist carries out the entire process, including chiral film construction, characterization, and testing. A machine learned reverse design model using spectrum embedded descriptors is developed to predict optimal structure/process parameters for any targeted chiroptical property. A series of chiral films with a dissymmetry factor as high as 1.9 ( g abs  ~ 1.9) are identified out of more than 100 million possible structures, and their feasible application in circular polarization-selective color filters for multiplex laser display and switchable circularly polarized (CP) luminescence is demonstrated. Our findings not only provide chiral films with the highest reported chiroptical activity, but also have great fundamental value for the inverse design of chiroptical materials. Chiral materials with strong and tunable chiroptical activities are highly desirable. Here, the authors employ an AI-guided robotic platform to fully execute a cyclic process to perform inverse design and fabricate chiral films with target chiroptical performance.

To address the fixed-parameter limitations of traditional PID control (e.g., excessive overshoot, prolonged settling time, poor adaptability to nonlinearities) and the insufficient real-time adjustment capability of conventional fuzzy PID control, which relies on empirically predefined rule bases, this study proposes a self-correcting fuzzy PID control strategy for agricultural water-fertilizer integrated systems. Traditional PID control, due to its static parameters, suffers from reduced stability and error accumulation under dynamic variations (e.g., irrigation flow fluctuations, environmental disturbances) or nonlinear interactions (e.g., coupling effects of fertilizer concentration and pH). While conventional fuzzy PID control incorporates fuzzy reasoning, its offline-designed rule bases and membership functions lack online adaptive parameter correction, leading to degraded precision in complex operating conditions. To tackle challenges posed by uncertain variables (e.g., time-varying soil permeability) and nonlinear parameters resistant to precise mathematical modeling, this research integrates fuzzy logic with an online self-correcting mechanism, constructs a mathematical model for the integrated control system, designs real-time correction rules, and validates the model through simulations using Matlab/Simulink and a semi-physical PC platform. The results demonstrate that the self-correcting fuzzy PID control significantly optimizes key performance metrics: overshoot (reduced by 21.3%), settling time (shortened by 34.7%), and steady-rate error (decreased by 18.9%), outperforming both traditional PID and fuzzy PID methods in concentration and pH regulation. Its parameter self-adaptation capability effectively balances dynamic response and steady-state performance, resolving issues such as overshoot oscillation and lagging regulation in nonlinear dynamics. In practical applications, the system achieved an average plant height growth rate of 15.86%-21.73% and a 30.41% yield improvement compared to the control group, validating the enhanced synergistic control of water and fertilizer enabled by the variable universe fuzzy PID approach. This study provides a robust control solution with theoretical innovation and practical value for managing complex nonlinear systems in precision agriculture.

Graph convolutional networks are widely used in skeleton-based action recognition because of their good fitting ability to non-Euclidean data. While conventional multi-scale temporal convolution uses several fixed-size convolution kernels or dilation rates at each layer of the network, we argue that different layers and datasets require different receptive fields. We use multi-scale adaptive convolution kernels and dilation rates to optimize traditional multi-scale temporal convolution with a simple and effective self attention mechanism, allowing different network layers to adaptively select convolution kernels of different sizes and dilation rates instead of being fixed and unchanged. Besides, the effective receptive field of the simple residual connection is not large, and there is a great deal of redundancy in the deep residual network, which will lead to the loss of context when aggregating spatio-temporal information. This article introduces a feature fusion mechanism that replaces the residual connection between initial features and temporal module outputs, effectively solving the problems of context aggregation and initial feature fusion. We propose a multi-modality adaptive feature fusion framework (MMAFF) to simultaneously increase the receptive field in both spatial and temporal dimensions. Concretely, we input the features extracted by the spatial module into the adaptive temporal fusion module to simultaneously extract multi-scale skeleton features in both spatial and temporal parts. In addition, based on the current multi-stream approach, we use the limb stream to uniformly process correlated data from multiple modalities. Extensive experiments show that our model obtains competitive results with state-of-the-art methods on the NTU-RGB+D 60 and NTU-RGB+D 120 datasets.

BackgroundRadioimmunotherapy has a promising antitumor effect in hepatocellular carcinoma (HCC), depending on the regulatory effect of radiotherapy on tumor immune microenvironment. Ionizing radiation (IR)-induced DNA damage repair (DDR) pathway activation leads to the inhibition of immune microenvironment, thus impairing the antitumor effect of radioimmunotherapy. However, it is unclear whether inhibition of the DDR pathway can enhance the effect of radioimmunotherapy. In this study, we aim to explore the role of DDR inhibitor AZD6738 on the combination of radiotherapy and immune checkpoint inhibitors (ICIs) in HCC.MethodsC57BL/6 mouse subcutaneous tumor model was used to evaluate the ability of different treatment regimens in tumor growth control and tumor recurrence inhibition. Effects of each treatment regimen on the alterations of immunophenotypes including the quantification, activation, proliferating ability, exhaustion marker expression, and memory status were assessed by flow cytometry.ResultsAZD6738 further increased radiotherapy-stimulated CD8+ T cell infiltration and activation and reverted the immunosuppressive effect of radiation on the number of Tregs in mice xenografts. Moreover, compared with radioimmunotherapy (radiotherapy plus anti-PD-L1 (Programmed death ligand 1)), the addition of AZD6738 boosted the infiltration, increased cell proliferation, enhanced interferon (IFN)-γ production ability of TIL (tumor-infiltrating lymphocyte) CD8+ T cells, and caused a decreasing trend in the number of TIL Tregs and exhausted T cells in mice xenografts. Thus, the tumor immune microenvironment was significantly improved. Meanwhile, triple therapy (AZD6738 plus radiotherapy plus anti-PD-L1) also induced a better immunophenotype than radioimmunotherapy in mice spleens. As a consequence, triple therapy displayed greater benefit in antitumor efficacy and mice survival than radioimmunotherapy. Mechanism study revealed that the synergistic antitumor effect of AZD6738 with radioimmunotherapy relied on the activation of cyclic GMP–AMP synthase /stimulator of interferon genes (cGAS/STING) signaling pathway. Furthermore, triple therapy led to stronger immunologic memory and lasting antitumor immunity than radioimmunotherapy, thus preventing tumor recurrence in mouse models.ConclusionsOur findings indicate that AZD6738 might be a potential synergistic treatment for radioimmunotherapy to control the proliferation of HCC cells, prolong survival, and prevent tumor recurrence in patients with HCC by improving the immune microenvironment.

Nonlinear friction in high-precision, ultra-low-speed servo systems severely degrades performance, causing low-speed crawling, static errors, and limit-cycle oscillations. This study introduces the LuGre friction model to describe these phenomena mathematically and proposes an improved genetic algorithm (GA) for precise parameter identification. Simulations demonstrate that LuGre-based feedforward compensation outperforms conventional proportional-integral-derivative (PID) control, effectively mitigating speed tracking errors and enhancing both speed and position accuracy. Experimental validation on a linear motor platform confirms the method’s efficacy, achieving a 25.1% improvement in tracking accuracy. The results highlight the practical relevance of this approach for precision servo systems. This work has achieved a practical identification framework for LuGre parameters, combining GA optimization with transient/steady-state data, feedforward compensation that directly injects estimated friction forces, bypassing feedback delays and experimental verification of the method’s industrial applicability.

The segmentation of the left ventricle (LV) wall in four-chamber view cardiac sequential image is significant for cardiac disease diagnosis and cardiac mechanisms study; however, there is no successful reported work on sequential four-chambered view LV wall segmentation due to the complex four-chamber structure and diversity of wall motion. In this article, we propose a dense recurrent neural network (RNN) algorithm to achieve accurately LV wall segmentation in a four-chamber view MRI time sequence. In the cardiac sequential LV wall process, not only the sequential accuracy but also the accuracy of each image matters. Thus, we propose a dense RNN to provide compensation for the first long short-term memory (LSTM) cells. Two RNNs are combined in this work, the first one aims at providing information for the first image, and the second RNN generates segmentation result. In this way, the proposed dense RNN improves the accuracy of the first frame image. What is more is that, it improves the effectiveness of information flow between LSTM cells. Obtaining more competent information from the former cell, frame-wise segmentation accuracy is greatly improved. Based on the segmentation result, an algorithm is proposed to estimate cardiac state. This is the first time that deals with both cardiac time-sequential LV segmentation problems and, robustly, estimates cardiac state. Rather than segmenting each frame separately, utilizing cardiac sequence information is more stable. The proposed method ensures an Intersection over Union (IoU) of 92.13%, which outperforms other classical deep learning algorithms.

A duplex surface enhanced Raman scattering (SERS)-based lateral flow immunosensor was established for the simultaneous detection of two common antibiotic residues including tetracycline and penicillin in milk. The newly synthesized Au@Ag nanoparticles were labeled with different Raman molecules including 5,5-dithiobis-2-nitrobenzoic acid (DTNB) or 4-mercaptobenzoic acid (MBA), followed by the conjugation of anti-tetracycline monoclonal antibody or anti-penicillin receptor, forming two kinds of SERS nanoprobes. The two nanoprobes can recognize tetracycline-BSA and ampicillin-BSA, respectively, which facilitates the simultaneous detection of the two types of antibiotics on a single test line. After optimization, detection limits of tetracycline and penicillin as low as 0.015 ng/mL and 0.010 ng/mL, respectively, were achieved. These values were far below those of most of other documented bio-analytical approaches. Moreover, the spiking test demonstrates an excellent assay accuracy with recoveries of 88.8% to 111.3%, and satisfactory assay precision with relative standard deviation below 16%. Consequently, the results demonstrate that the SERS-based lateral flow immunosensor developed in this study has the advantages of excellent assay sensitivity and remarkable multiplexing capability, thus it will have great application potential in food safety monitoring.

Perennial ryegrass ( ) is one of the most widely used forage and turf grasses in the world due to its desirable agronomic qualities. However, as a cool-season perennial grass species, high temperature is a major factor limiting its performance in warmer and transition regions. In this study, a transcriptome was generated using a cDNA library constructed from perennial ryegrass leaves subjected to short-term heat stress treatment. Then the expression profiling and identification of perennial ryegrass heat response genes by digital gene expression analyses was performed. The goal of this work was to produce expression profiles of high temperature stress responsive genes in perennial ryegrass leaves and further identify the potentially important candidate genes with altered levels of transcript, such as those genes involved in transcriptional regulation, antioxidant responses, plant hormones and signal transduction, and cellular metabolism. The assembly of perennial ryegrass transcriptome in this study obtained more total and annotated unigenes compared to previously published ones. Many DEGs identified were genes that are known to respond to heat stress in plants, including HSFs, HSPs, and antioxidant related genes. In the meanwhile, we also identified four gene candidates mainly involved in C carbon fixation, and one TOR gene. Their exact roles in plant heat stress response need to dissect further. This study would be important by providing the gene resources for improving heat stress tolerance in both perennial ryegrass and other cool-season perennial grass plants.