Interaction-knowledge semantic alignment for recommendation.

In order to alleviate the issue of data sparsity, knowledge graphs are introduced into recommender systems because they contain diverse information about items. The existing knowledge graph enhanced recommender systems utilize both user-item interaction data and knowledge graph, but those methods ignore the semantic difference between interaction data and knowledge graph. On the other hand, for the item representations obtained from two kinds of graph structure data respectively, the existing methods of fusing representations only consider the item representations themselves, without considering the personalized preference of users. In order to overcome the limitations mentioned above, we present a recommendation method named Interaction-Knowledge Semantic Alignment for Recommendation (IKSAR). By introducing a semantic alignment module, the semantic difference between the interaction bipartite graph and the knowledge graph is reduced. The representation of user is integrated during the fusion of representations of item, which improves the quality of the fused representation of item. To validate the efficacy of the proposed approach, we perform comprehensive experiments on three datasets. The experimental results demonstrate that the IKSAR is superior to the existing methods, showcasing notable improvement.

Study on the source identification and migration and transformation characteristics of heavy metal pollution in sediments of Chaohu Lake Tongyang River Basin.

Heavy metal pollution in soil and groundwater is difficult to detect in time because of its long-term accumulation. This poses a severe hazard to the ecosystem security in polluted areas. Considering the Tongyang River Basin of the Chaohu Lake water system as the research object, various pollution evaluation methods, Geographic Information System (GIS) source analysis models, adsorption and desorption experiments, and mathematical models were used. Based on the data of the heavy metal content in the soil and river sediments in the study area, regional heavy metal pollution evaluation, spatial distribution characteristics, source identification, and lake loss analysis were conducted to evaluate the current situation of environmental pollution, pollution sources, and migration and transformation of heavy metals in the basin. The results show that the heavy metals in the soil of the basin are mainly distributed from the southwest to the northeast and the island distribution to the east of Tongyang Town. The Cd, Cu, Zn, and Pb concentrations are higher in the sediment than in the soil. The Unmix analytical model was established. The natural source of heavy metals in the Tongyang River Basin and the composite pollution source of traffic-production and life showed a high correlation. According to the one-dimensional steady-state model, the concentrations of the heavy metals are in the following descending order: Zn, Cr, Pb, Cu, Ni, and Cd. The amounts of these elements entering the lake annually are 1993.04, 1028.1, 372.47, 301.72, 308.80, and 19.46 kg, respectively (a total of 4023.6 kg).

Development of highly adaptable RT-PCR methods for identifying Delta and BA.1 variants in inactivated COVID-19 vaccines.

Background The emergence and rapid spread of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), poses a significant threat to human health and public safety. While next-generation sequencing (NGS) is capable of detecting and tracking new COVID-19 variants for disease diagnosis and prevention, its high cost and time-consuming nature limit its widespread use. In this study, our aim was to develop a highly adaptable and accurate RT-PCR method for identifying the Delta or BA.1 variants in inactivated COVID-19 vaccine. We devised three two-plex RT-PCR methods targeting specific mutation sites: S: Δ156-157, S: N211-, L212I, and S: Δ142-144, Y145D. The RT-PCR method targeting the S: Δ156-157 mutation site was able to distinguish the Delta variant from other COVID-19 virus strains, while the RT-PCR methods targeting the S: N211-, L212I or S: Δ142-144, Y145D mutation sites were able to distinguish the BA.1 variant from other COVID-19 virus strains. We separately validated these three two-plex RT-PCR methods, and the results demonstrated good linearity, repeatability, reproducibility, and specificity for each method. Moreover, all three methods can be applied in the production of SARS-CoV-2 variant inactivated vaccines, enabling the identification of Delta or BA.1 variants in virus cultures as well as in inactivated vaccine stocks. This study presents a systematic approach to identify COVID-19 variants using multiple RT-PCR methods. We successfully developed three two-plex RT-PCR methods that can identify Delta and BA.1 variants based on specific mutation sites, and we completed the validation of these three methods.

Adverse Associations of Long-Term Exposure to PM2.5 and Its Components with Platelet Traits among Subway Shift-Workers without Air Purifier Use.

Air purifier use, shift work, and long-term exposure to fine particulate matter (PM2.5) are linked to platelet abnormality. However, the role of air purifier use and shift work in the individual or joint associations of PM2.5 and its components with platelet indices are largely unknown. A total of 8772 participants were recruited from a population of subway workers in China. PM2.5 and its component data were obtained from the Tracking Air Pollution in China dataset. The role of air purifier use and shift work in the association between PM2.5 and its components and platelet indices were analyzed. Among shift workers without air purifier use, positive associations of PM2.5 and each component in PM2.5 with the mean platelet volume (MPV) or platelet counts (PLT) were observed, whereas negative associations of PM2.5 and each component in PM2.5 with the platelet distribution width (PDW) were observed. Furthermore, estimated changes (95%CIs) in PLT, MPV, and PDW in response to each 10th percentile increment in the mixture of PM2.5 and its components were 0.8657 (0.2496, 1.4819), 0.0192 (0.0054, 0.0329), and -0.0648 (-0.0945, -0.0351), respectively, and sulfate in PM2.5 was the major contributor to those associations. Long-term exposure to PM2.5 and its components was related to increased platelet disorders among shift workers without air purifier use, and those associations were mainly attributed to sulfate in PM2.5.

Induction-concurrent chemoradiotherapy with or without sintilimab in patients with locoregionally advanced nasopharyngeal carcinoma in China (CONTINUUM): a multicentre, open-label, parallel-group, randomised, controlled, phase 3 trial.

BACKGROUND: Anti-PD-1 therapy and chemotherapy is a recommended first-line treatment for recurrent or metastatic nasopharyngeal carcinoma, but the role of PD-1 blockade remains unknown in patients with locoregionally advanced nasopharyngeal carcinoma. We assessed the addition of sintilimab, a PD-1 inhibitor, to standard chemoradiotherapy in this patient population. METHODS: This multicentre, open-label, parallel-group, randomised, controlled, phase 3 trial was conducted at nine hospitals in China. Adults aged 18-65 years with newly diagnosed high-risk non-metastatic stage III-IVa locoregionally advanced nasopharyngeal carcinoma (excluding T3-4N0 and T3N1) were eligible. Patients were randomly assigned (1:1) using blocks of four to receive gemcitabine and cisplatin induction chemotherapy followed by concurrent cisplatin radiotherapy (standard therapy group) or standard therapy with 200 mg sintilimab intravenously once every 3 weeks for 12 cycles (comprising three induction, three concurrent, and six adjuvant cycles to radiotherapy; sintilimab group). The primary endpoint was event-free survival from randomisation to disease recurrence (locoregional or distant) or death from any cause in the intention-to-treat population. Secondary endpoints included adverse events. This trial is registered with ClinicalTrials.gov (NCT03700476) and is now completed; follow-up is ongoing. FINDINGS: Between Dec 21, 2018, and March 31, 2020, 425 patients were enrolled and randomly assigned to the sintilimab (n=210) or standard therapy groups (n=215). At median follow-up of 41·9 months (IQR 38·0-44·8; 389 alive at primary data cutoff [Feb 28, 2023] and 366 [94%] had at least 36 months of follow-up), event-free survival was higher in the sintilimab group compared with the standard therapy group (36-month rates 86% [95% CI 81-90] vs 76% [70-81]; stratified hazard ratio 0·59 [0·38-0·92]; p=0·019). Grade 3-4 adverse events occurred in 155 (74%) in the sintilimab group versus 140 (65%) in the standard therapy group, with the most common being stomatitis (68 [33%] vs 64 [30%]), leukopenia (54 [26%] vs 48 [22%]), and neutropenia (50 [24%] vs 46 [21%]). Two (1%) patients died in the sintilimab group (both considered to be immune-related) and one (<1%) in the standard therapy group. Grade 3-4 immune-related adverse events occurred in 20 (10%) patients in the sintilimab group. INTERPRETATION: Addition of sintilimab to chemoradiotherapy improved event-free survival, albeit with higher but manageable adverse events. Longer follow-up is necessary to determine whether this regimen can be considered as the standard of care for patients with high-risk locoregionally advanced nasopharyngeal carcinoma. FUNDING: National Natural Science Foundation of China, Key-Area Research and Development Program of Guangdong Province, Natural Science Foundation of Guangdong Province, Overseas Expertise Introduction Project for Discipline Innovation, Guangzhou Municipal Health Commission, and Cancer Innovative Research Program of Sun Yat-sen University Cancer Center. TRANSLATION: For the Chinese translation of the abstract see Supplementary Materials section.

Surveillance for rhabdomyolysis after the consumption of crayfish in Wuhan, China, 2016-2022.

Objectives: To analyze the epidemiological characteristics and etiology of crayfish-related rhabdomyolysis. Methods: Cases of crayfish-related rhabdomyolysis in Wuhan were monitored, and professional training of city's surveillance personnel was conducted. Unified questionnaires were used to collect data. Results: The first case of crayfish-related rhabdomyolysis occurred on July 12, 2016. Subsequently, 423 patients were reported over the next 7 years, with muscle pain, weakness, and chest distress as main symptoms. In total, 64.54% (273/423) of patients were females, and young adults (aged 20-49 years) account for 86.22% (363/423) of patients. The primary clinical presentations were muscle pain, muscle weakness, and chest discomfort. The median incubation time was 6 h. And the number of cases may be related to water levels in Yangzi river. Laboratory tests revealed elevated creatine kinase and myoglobin levels. In total, 95.16% (236/248) of patients had consumed crayfish tail shrimp and 91.53% (227/248) had consumed crayfish liver and pancreas (Female crayfish also contain ovaries). Only 25.00% (62/248) of patients had a history of alcohol consumption. On average, 227 patients consumed 15 (3-50) crayfish, of whom 84.14% (191/227) consumed more than 10 crayfish. All patients had a favorable prognosis. Conclusion: Crayfish-related rhabdomyolysis is a kind of a case or cluster of patients present with severe myalgia or weakness of unknown etiology and mechanism disease in Wuhan, China, 2016-2022. Excessive consumption of crayfish may be a risk factor for the disease. The relationship between the specific parts of crayfish consumed and the onset of the disease is unclear, suggesting further research is needed to identify the relevant risk factors for the disease.

Does the Dose of Standard Adjuvant Chemotherapy Affect the Triple-negative Breast Cancer Benefit from Extended Capecitabine Metronomic Therapy? An Exploratory Analysis of the SYSUCC-001 Trial.

Purpose: Results from studies of extended capecitabine after the standard adjuvant chemotherapy in early stage triple-negative breast cancer (TNBC) were inconsistent, and only low-dose capecitabine from the SYSUCC-001 trial improved disease-free survival (DFS). Adjustment of the conventional adjuvant chemotherapy doses affect the prognosis and may affect the efficacy of subsequent treatments. This study investigated whether the survival benefit of the SYSUCC-001 trial was affected by dose adjustment of the standard adjuvant chemotherapy or not. Patients and Methods: We reviewed the adjuvant chemotherapy regimens before the extended capecitabine in the SYSUCC-001 trial. Patients were classified into "consistent" (standard acceptable dose) and "inconsistent" (doses lower than acceptable dose) dose based on the minimum acceptable dose range in the landmark clinical trials. Cox proportional hazards model was used to investigate the impact of dose on the survival outcomes. Results: All 434 patients in SYSUCC-001 trial were enrolled in this study. Most of patients administered the anthracycline-taxane regimen accounted for 88.94%. Among patients in the "inconsistent" dose, 60.8% and 47% received lower doses of anthracycline and taxane separately. In the observation group, the "inconsistent" dose of anthracycline and taxane did not affect DFS compared with the "consistent" dose. Moreover, in the capecitabine group, the "inconsistent" anthracycline dose did not affect DFS compared with the "consistent" dose. However, patients with "consistent" taxane doses benefited significantly from extended capecitabine (P=0.014). The sufficient dose of adjuvant taxane had a positive effect of extended capecitabine (hazard ratio [HR] 2.04; 95% confidence interval [CI] 1.02 to 4.06). Conclusion: This study found the dose reduction of adjuvant taxane might negatively impact the efficacy of capecitabine. Therefore, the reduction of anthracycline dose over paclitaxel should be given priority during conventional adjuvant chemotherapy, if patients need dose reduction and plan for extended capecitabine.

Adaptation process of decellularized vascular grafts as hemodialysis access in vivo.

Arteriovenous grafts (AVGs) have emerged as the preferred option for constructing hemodialysis access in numerous patients. Clinical trials have demonstrated that decellularized vascular graft exhibits superior patency and excellent biocompatibility compared to polymer materials; however, it still faces challenges such as intimal hyperplasia and luminal dilation. The absence of suitable animal models hinders our ability to describe and explain the pathological phenomena above and in vivo adaptation process of decellularized vascular graft at the molecular level. In this study, we first collected clinical samples from patients who underwent the construction of dialysis access using allogeneic decellularized vascular graft, and evaluated their histological features and immune cell infiltration status 5 years post-transplantation. Prior to the surgery, we assessed the patency and intimal hyperplasia of the decellularized vascular graft using non-invasive ultrasound. Subsequently, in order to investigate the in vivo adaptation of decellularized vascular grafts in an animal model, we attempted to construct an AVG model using decellularized vascular grafts in a small animal model. We employed a physical-chemical-biological approach to decellularize the rat carotid artery, and histological evaluation demonstrated the successful removal of cellular and antigenic components while preserving extracellular matrix constituents such as elastic fibers and collagen fibers. Based on these results, we designed and constructed the first allogeneic decellularized rat carotid artery AVG model, which exhibited excellent patency and closely resembled clinical characteristics. Using this animal model, we provided a preliminary description of the histological features and partial immune cell infiltration in decellularized vascular grafts at various time points, including Day 7, Day 21, Day 42, and up to one-year post-implantation. These findings establish a foundation for further investigation into the in vivo adaptation process of decellularized vascular grafts in small animal model.

Reliability-Guided Hierarchical Memory Network for Scribble-Supervised Video Object Segmentation.

This article aims to solve the video object segmentation (VOS) task in a scribble-supervised manner, in which VOS models are not only initialized with sparse target scribbles for inference but also trained by sparse scribble annotations. Thus, the annotation burdens for both initialization and training can be substantially lightened. The difficulties of scribble-supervised VOS lie in two aspects: 1) it demands a strong reasoning ability to carefully segment the target given only a sparse initial target scribble and 2) it necessitates learning dense prediction from sparse scribble annotations during training, requiring powerful learning capability. In this work, we propose a reliability-guided hierarchical memory network (RHMNet) for this task, which segments the target in a stepwise expanding strategy w.r.t. the memory reliability level. To be specific, RHMNet maintains a reliability-guided memory bank. It first uses the high-reliability memory to locate the region with high reliability belonging to the target, i.e., highly similar to the initial target scribble. Then, it expands the located high-reliability region to the entire target conditioned on the region itself and all existing memories. In addition, we propose a scribble-supervised learning mechanism to facilitate the model learning for dense prediction. It exploits the pixel-level relations within a single frame and the instance-level variations across multiple frames to take full advantage of the scribble annotations in sequence training samples. The favorable performance on four popular benchmarks demonstrates that our method is promising. Our project is available at: https://github.com/mkg1204/RHMNet-for-SSVOS.

Immune-Related Genes and Immune Cell Infiltration Characterize the Maturation Status of Arteriovenous Fistulas: An Integrative Bioinformatics Study and Experimental Validation Based on Transcriptome Sequencing.

Purpose: Arteriovenous fistula (AVF) is the preferred vascular access for hemodialysis, but the low maturation rate is concerning. Immune cells' impact on AVF maturation lacks bioinformatics research. The study aims to investigate the potential predictive role of immune-related genes and immune cell infiltration characteristics in AVF maturation. Patients and Methods: We analyzed the high-throughput sequencing dataset to identify differentially expressed genes (DEGs). Then, we performed enrichment analyses (GO, KEGG, GSEA) on immune-related genes and pathways in mature AVF. We focused on differentially expressed immune-related genes (DEIRGs) and constructed a PPI network to identify hub genes. These hub genes were validated in other databases and experiments, including qPCR and immunohistochemistry (IHC). The immune cell infiltration characteristics in native veins, failed AVFs, and matured AVFs were analyzed by cibersortX. Partial experimental validation was conducted using clinical samples. Results: Our results showed that immune-related genes and signaling pathways are significantly enriched in mature AVF. We validated this in other databases and ultimately identified three hub genes (IL1B, IL6, CXCR4) in combination with experiments. Significant differences in immune cell infiltration characteristics were observed among native veins, failed AVFs, and matured AVFs. Immune cell infiltration analysis revealed that accumulation of CD4+ T cells, dendritic cells, mast cells and M2 macrophages contribute to AVF maturation. These immune-related genes and immune cells have the potential to serve as predictive factors for AVF maturation. We partially validated this experimentally. Conclusion: From a bioinformatics perspective, our results have identified, for the first time, a set of immune-related genes and immune cell infiltration features that can characterize the maturation of AVF and significantly impact AVF maturation. These features hold potential as predictive indicators for AVF maturation outcomes.

Safety, pharmacokinetics and pharmacodynamics of HWH486 capsules in healthy adults: A randomized, double-blind, placebo-controlled, phase I dose-escalation study.

OBJECTIVES: HWH486 inhibits Bruton's tyrosine kinase and therefore shows promise as a treatment against rheumatoid arthritis and chronic spontaneous urticaria. This phase I trial assessed tolerability, safety, pharmacokinetics and pharmacodynamics of a single oral dose of HWH486 capsules in healthy adults. METHODS: A single-center, randomized, double-blind, placebo-controlled, dose-escalation study from 10 to 800 mg was conducted in 96 healthy Chinese adults, of whom 80 received HWH486 and 16 received placebo. RESULTS: A total of 96 subjects were enrolled, and all completed the study. In the HWH486 group, mean Tmax ranged from 1.03 to 2.00 h, and mean T1/2 ranged from 0.85 to 8.67 h across the dose range from 10 to 800 mg. Mean Cmax increased linearly with dose, while mean AUC0-t increased non-linearly. Occupancy of Bruton's tyrosine kinase peaked within 0.50-4.00 h after administration across the dose groups, and the delay until peak occupancy decreased with increasing dose. Twenty-five subjects (31.25 %) in the HWH486 group experienced 35 treatment-emergent adverse events, while four subjects (25.00 %) in the placebo group experienced eight such events. CONCLUSIONS: HWH486 is well tolerated and safe in healthy adults, in whom it can strongly bind Bruton's tyrosine kinase. These findings justify clinical studies of HWH486 efficacy against autoimmune diseases.

Nomogram Update to Predict the High Genomic Risk Breast Cancer by Different Races.

PURPOSE: To develop a nomogram to predict the high-risk recurrence score (RS) and to customize the nomogram for different races in early-stage hormone receptor (HoR)-positive, human epidermal growth factor receptor-2 (HER2)-negative breast cancer. METHODS: Patients diagnosed between 2010 and 2015 were included from the surveillance, epidemiology, and end results oncotype DX database. The nomogram was assessed with a receiver operating characteristic curve to measure the area under the curve (AUC) with a 95% confidence interval (95% CI). The nomogram was developed and internally validated for discrimination and calibration, and then validated in different races. RESULTS: A total of 48,464 patients were included and randomly assigned to the training cohort (n = 36370, 75.0%) and validation cohort (n = 12,094, 25.0%). Patients in the training cohort were identified to develop the nomogram, including 32,683 (89.9%) White women, 3135 (8.6%) Black women, and 552 (1.5%) Chinese women. Five independent predictive factors for high-risk RS were included to develop the nomogram, including tumor grade, progesterone receptor status, histological subtype, race, and tumor stage. The AUC was 0.696 (95% CI, 0.682-0.710) in the training cohort and 0.700 (95% CI, 0.676-0.724) in the validation cohort. There was no significant difference between the training cohort and the validation cohort. When validating the nomogram classified by race, the AUC was 0.694 (95% CI, 0.682-0.706) for the White cohort, 0.708 (95% CI, 0.673-0.743) for the Black cohort, and 0.653 (95% CI, 0.565-0.741) for the Chinese cohort. CONCLUSION: The developed nomogram for predicting high-risk RS is available for different races in patients with HoR+/HER2- breast cancer, which could be used as qualified surrogates before ordering the 21-gene RS testing.

Exposure to multiple neonicotinoid insecticides, oxidative stress, and gestational diabetes mellitus: Association and potential mediation analyses.

As the most extensively used insecticides worldwide, neonicotinoid insecticides (NNIs) have received a growing global concern over their adverse health effects. This study aimed to assess the associations of urinary concentrations of NNIs in early pregnancy with gestational diabetes mellitus (GDM) and the mediation roles of oxidative DNA damage, RNA damage, and lipid peroxidation in the associations. With a prospective nested case-control study, 519 GDM cases and 519 controls were matched on the infant's sex and maternal age. Urinary biomarkers of NNIs exposure and oxidative stress were measured in early pregnancy. We estimated the associations of single and the mixture of NNIs and their metabolites with GDM by conditional logistic regression and quantile g-computation models, respectively. The mediating roles of oxidative stress were evaluated by the structural equation model. The odds of GDM significantly increased by 15 %, 18 %, 26 %, 42 %, 49 %, and 13 % in each unit increment of ln-transformed concentrations of urinary imidacloprid (IMI), imidacloprid-olefin (IMI-olefin), desnitro-imidacloprid (DN-IMI), thiamethoxam (THM), clothianidin, and desmethyl-clothianidin, respectively. Exposure to the mixture of NNIs was associated with increased odds of GDM (adjusted OR: 1.76; 95 %CI: 1.45, 2.13). Advanced maternal age enhanced the associations of 5-hydroxy-IMI, DN-IMI, and IMI-olefin with GDM (P < 0.05), and being overweight/obese before pregnancy strengthened the effects of IMI, IMI-olefin, and THM on GDM (P < 0.05). In the association of NNIs exposure and GDM, the proportions mediated by oxidative DNA damage, RNA damage, and overall oxidative stress were 9.8 %, 11.8 %, and 14.5 %, respectively (P < 0.05). Exposure to individual NNIs and a mixture of NNIs were associated with GDM, and maternal age and pre-pregnancy BMI may modify the association. The possible mechanism underlying the association between NNIs and GDM may involve oxidative damage to nucleic acids.

Association between Nutrition and Health Knowledge and Multiple Chronic Diseases: A Large Cross-Sectional Study in Wuhan, China.

Nutrition and health knowledge (NHK) is linked to people's dietary behavior and health outcomes. However, studies on the associations between NHK and chronic diseases are limited. This study aimed to examine the association of NHK with five specific chronic diseases (diabetes/hyperglycemia, hypertension, dyslipidemia, coronary heart disease (CHD), and stroke) in central China. Individual NHK and disease status were investigated using a self-reporting questionnaire. We further added up the number of chronic diseases and used this as a secondary outcome. A total of 21,559 adults were enrolled in this cross-sectional study. NHK score was significantly inversely associated with diabetes/hyperglycemia, hypertension, CHD, and stroke (all p-trends < 0.001). Moreover, an inverse association was found between NHK and the number of chronic diseases, especially among responders with three or more chronic diseases. Stratified analyses showed that the above association was more likely to be stronger among younger, female, highly educated, and inner-city residents. However, NHK was negatively associated with dyslipidemia in less educated people and positively correlated with dyslipidemia in highly educated people. NHK showed an inverse relationship with specific chronic diseases and the number of chronic diseases. Improving NHK might be a key strategy for easing the global burden of chronic diseases.

CircNCOR1 regulates breast cancer radiotherapy efficacy by regulating CDK2 via hsa-miR-638 binding.

BACKGROUND: Despite aggressive local and regional therapy, triple-negative breast cancer (TNBC) is characterized by an increased risk of locoregional recurrence. RNA-sequencing data has identified a large number of circRNAs in primary breast cancers, but the role of specific circRNAs in regulating the radiosensitivity of TNBC is not fully understood. This research aimed to investigate the function of circNCOR1 in the radiosensitivity of TNBC. METHODS: CircRNA high-throughput sequencing was conducted on two breast cancer MDA-MB-231 and BT549 cell lines after 6 Gy radiation. The relationship between circNCOR1, hsa-miR-638, and CDK2 was determined by RNA immunoprecipitation (RIP), FISH and luciferase assays. The proliferation and apoptosis of breast cancer cells were measured by CCK8, flow cytometry, colony formation assays, and western blot. RESULTS: Differential expression of circRNAs was closely related to the proliferation of breast cancer cells after irradiation. Overexpression of circNCOR1 facilitated the proliferation of MDA-MB-231 and BT549 cells and impaired the radiosensitivity of breast cancer cells. Additionally, circNCOR1 acted as a sponge for hsa-miR-638 to regulate the downstream target protein CDK2. Overexpression of hsa-miR-638 promoted apoptosis of breast cancer cells, while overexpression of CDK2 alleviated apoptosis and increased proliferation and clonogenicity. In vivo, overexpression of circNCOR1 partially reversed radiation-induced loosening of tumor structures and enhanced tumor cell proliferation. CONCLUSION: Our results demonstrated that circNCOR1 bounds to hsa-miR-638 and targets CDK2, thereby regulating the radiosensitivity of TNBC.

Audio2Gestures: Generating Diverse Gestures From Audio.

People may perform diverse gestures affected by various mental and physical factors when speaking the same sentences. This inherent one-to-many relationship makes co-speech gesture generation from audio particularly challenging. Conventional CNNs/RNNs assume one-to-one mapping, and thus tend to predict the average of all possible target motions, easily resulting in plain/boring motions during inference. So we propose to explicitly model the one-to-many audio-to-motion mapping by splitting the cross-modal latent code into shared code and motion-specific code. The shared code is expected to be responsible for the motion component that is more correlated to the audio while the motion-specific code is expected to capture diverse motion information that is more independent of the audio. However, splitting the latent code into two parts poses extra training difficulties. Several crucial training losses/strategies, including relaxed motion loss, bicycle constraint, and diversity loss, are designed to better train the VAE. Experiments on both 3D and 2D motion datasets verify that our method generates more realistic and diverse motions than previous state-of-the-art methods, quantitatively and qualitatively. Besides, our formulation is compatible with discrete cosine transformation (DCT) modeling and other popular backbones (i.e. RNN, Transformer). As for motion losses and quantitative motion evaluation, we find structured losses/metrics (e.g. STFT) that consider temporal and/or spatial context complement the most commonly used point-wise losses (e.g. PCK), resulting in better motion dynamics and more nuanced motion details. Finally, we demonstrate that our method can be readily used to generate motion sequences with user-specified motion clips on the timeline.

Active Learning for Deep Visual Tracking.

Convolutional neural networks (CNNs) have been successfully applied to the single target tracking task in recent years. Generally, training a deep CNN model requires numerous labeled training samples, and the number and quality of these samples directly affect the representational capability of the trained model. However, this approach is restrictive in practice, because manually labeling such a large number of training samples is time-consuming and prohibitively expensive. In this article, we propose an active learning method for deep visual tracking, which selects and annotates the unlabeled samples to train the deep CNN model. Under the guidance of active learning, the tracker based on the trained deep CNN model can achieve competitive tracking performance while reducing the labeling cost. More specifically, to ensure the diversity of selected samples, we propose an active learning method based on multiframe collaboration to select those training samples that should be and need to be annotated. Meanwhile, considering the representativeness of these selected samples, we adopt a nearest-neighbor discrimination method based on the average nearest-neighbor distance to screen isolated samples and low-quality samples. Therefore, the training samples' subset selected based on our method requires only a given budget to maintain the diversity and representativeness of the entire sample set. Furthermore, we adopt a Tversky loss to improve the bounding box estimation of our tracker, which can ensure that the tracker achieves more accurate target states. Extensive experimental results confirm that our active-learning-based tracker (ALT) achieves competitive tracking accuracy and speed compared with state-of-the-art trackers on the seven most challenging evaluation benchmarks. Project website: https://sites.google.com/view/altrack/.

Self-Supervised Tracking via Target-Aware Data Synthesis.

While deep-learning-based tracking methods have achieved substantial progress, they entail large-scale and high-quality annotated data for sufficient training. To eliminate expensive and exhaustive annotation, we study self-supervised (SS) learning for visual tracking. In this work, we develop the crop-transform-paste operation, which is able to synthesize sufficient training data by simulating various appearance variations during tracking, including appearance variations of objects and background interference. Since the target state is known in all synthesized data, existing deep trackers can be trained in routine ways using the synthesized data without human annotation. The proposed target-aware data-synthesis method adapts existing tracking approaches within a SS learning framework without algorithmic changes. Thus, the proposed SS learning mechanism can be seamlessly integrated into existing tracking frameworks to perform training. Extensive experiments show that our method: 1) achieves favorable performance against supervised (Su) learning schemes under the cases with limited annotations; 2) helps deal with various tracking challenges such as object deformation, occlusion (OCC), or background clutter (BC) due to its manipulability; 3) performs favorably against the state-of-the-art unsupervised tracking methods; and 4) boosts the performance of various state-of-the-art Su learning frameworks, including SiamRPN++, DiMP, and TransT.

Exploring the Temporal Consistency of Arbitrary Style Transfer: A Channelwise Perspective.

Arbitrary image stylization by neural networks has become a popular topic, and video stylization is attracting more attention as an extension of image stylization. However, when image stylization methods are applied to videos, unsatisfactory results that suffer from severe flickering effects appear. In this article, we conducted a detailed and comprehensive analysis of the cause of such flickering effects. Systematic comparisons among typical neural style transfer approaches show that the feature migration modules for state-of-the-art (SOTA) learning systems are ill-conditioned and could lead to a channelwise misalignment between the input content representations and the generated frames. Unlike traditional methods that relieve the misalignment via additional optical flow constraints or regularization modules, we focus on keeping the temporal consistency by aligning each output frame with the input frame. To this end, we propose a simple yet efficient multichannel correlation network (MCCNet), to ensure that output frames are directly aligned with inputs in the hidden feature space while maintaining the desired style patterns. An inner channel similarity loss is adopted to eliminate side effects caused by the absence of nonlinear operations such as softmax for strict alignment. Furthermore, to improve the performance of MCCNet under complex light conditions, we introduce an illumination loss during training. Qualitative and quantitative evaluations demonstrate that MCCNet performs well in arbitrary video and image style transfer tasks. Code is available at https://github.com/kongxiuxiu/MCCNetV2.

LSOTB-TIR: A Large-Scale High-Diversity Thermal Infrared Single Object Tracking Benchmark.

Unlike visual object tracking, thermal infrared (TIR) object tracking methods can track the target of interest in poor visibility such as rain, snow, and fog, or even in total darkness. This feature brings a wide range of application prospects for TIR object-tracking methods. However, this field lacks a unified and large-scale training and evaluation benchmark, which has severely hindered its development. To this end, we present a large-scale and high-diversity unified TIR single object tracking benchmark, called LSOTB-TIR, which consists of a tracking evaluation dataset and a general training dataset with a total of 1416 TIR sequences and more than 643 K frames. We annotate the bounding box of objects in every frame of all sequences and generate over 770 K bounding boxes in total. To the best of our knowledge, LSOTB-TIR is the largest and most diverse TIR object tracking benchmark to date. We spilt the evaluation dataset into a short-term tracking subset and a long-term tracking subset to evaluate trackers using different paradigms. What's more, to evaluate a tracker on different attributes, we also define four scenario attributes and 12 challenge attributes in the short-term tracking evaluation subset. By releasing LSOTB-TIR, we encourage the community to develop deep learning-based TIR trackers and evaluate them fairly and comprehensively. We evaluate and analyze 40 trackers on LSOTB-TIR to provide a series of baselines and give some insights and future research directions in TIR object tracking. Furthermore, we retrain several representative deep trackers on LSOTB-TIR, and their results demonstrate that the proposed training dataset significantly improves the performance of deep TIR trackers. Codes and dataset are available at https://github.com/QiaoLiuHit/LSOTB-TIR.