Search

Modern Artificial Intelligence (AI) technologies, led by Machine Learning (ML), have gained unprecedented momentum over the past decade. Following this wave of ``AI summer", the network research community has also embraced AI/ML algorithms to address many problems related to network operations and management. However, compared to their counterparts in other domains, most ML-based solutions have yet to receive large-scale deployment due to insufficient maturity for production settings. This paper concentrates on the practical issues of developing and operating ML-based solutions in real networks. Specifically, we enumerate the key factors hindering the integration of AI/ML in real networks and review existing solutions to uncover the missing considerations. We also highlight two potential directions, i.e., MLOps and Causal ML, that can close the gap. We believe this paper spotlights the system-related considerations on implementing \& maintaining ML-based solutions and invigorate their full adoption in future networks., This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Local feature matching is challenging due to textureless and repetitive patterns. Existing methods focus on using appearance features and global interaction and matching, while the importance of geometry priors in local feature matching has not been fully exploited. Different from these methods, in this paper, we delve into the importance of geometry prior and propose Structured Epipolar Matcher (SEM) for local feature matching, which can leverage the geometric information in an iterative matching way. The proposed model enjoys several merits. First, our proposed Structured Feature Extractor can model the relative positional relationship between pixels and high-confidence anchor points. Second, our proposed Epipolar Attention and Matching can filter out irrelevant areas by utilizing the epipolar constraint. Extensive experimental results on five standard benchmarks demonstrate the superior performance of our SEM compared to state-of-the-art methods. Project page: https://sem2023.github.io., Comment: Accepted to CVPR Workshop 2023 (Image Matching: Local Features & Beyond). Project Page: https://sem2023.github.io

Weakly-supervised temporal action localization aims to localize and recognize actions in untrimmed videos with only video-level category labels during training. Without instance-level annotations, most existing methods follow the Segment-based Multiple Instance Learning (S-MIL) framework, where the predictions of segments are supervised by the labels of videos. However, the objective for acquiring segment-level scores during training is not consistent with the target for acquiring proposal-level scores during testing, leading to suboptimal results. To deal with this problem, we propose a novel Proposal-based Multiple Instance Learning (P-MIL) framework that directly classifies the candidate proposals in both the training and testing stages, which includes three key designs: 1) a surrounding contrastive feature extraction module to suppress the discriminative short proposals by considering the surrounding contrastive information, 2) a proposal completeness evaluation module to inhibit the low-quality proposals with the guidance of the completeness pseudo labels, and 3) an instance-level rank consistency loss to achieve robust detection by leveraging the complementarity of RGB and FLOW modalities. Extensive experimental results on two challenging benchmarks including THUMOS14 and ActivityNet demonstrate the superior performance of our method., Comment: Accepted by CVPR 2023. Code is available at https://github.com/RenHuan1999/CVPR2023_P-MIL

ObjectivesThis study aims to evaluate the diagnostic performance of machine-learning-based contrast-enhanced CT radiomic analysis for categorizing benign and malignant ovarian tumors.MethodsA total of 1,329 patients with ovarian tumors were randomly divided into a training cohort (N=930) and a validation cohort (N=399). All tumors were resected, and pathological findings were confirmed. Radiomic features were extracted from the portal venous phase images of contrast-enhanced CT. The clinical predictors included age, CA-125, HE-4, ascites, and margin of tumor. Both radiomics model (including selected radiomic features) and mixed model (incorporating selected radiomic features and clinical predictors) were constructed respectively. Six classifiers [k-nearest neighbor (KNN), support vector machines (SVM), random forest (RF), logistic regression (LR), multi-layer perceptron (MLP), and eXtreme Gradient Boosting (XGBoost)] were used for each model. The mean relative standard deviation (RSD) and area under the receiver operating characteristic curve (AUC) were applied to evaluate and select the best classifiers. Then, the performances of the two models with selected classifiers were assessed in the validation cohort.ResultsThe MLP classifier with the least RSD (1.21 and 0.53, respectively) was selected as the best classifier in both radiomics and mixed models. The two models with MLP classifier performed well in the validation cohort, with the AUCs of 0.91 and 0.96 and with accuracies (ACCs) of 0.83 and 0.87, respectively. The Delong test showed that the AUC of mixed model was statistically different from that of radiomics model (pConclusionsMachine-learning-based CT radiomic analysis could categorize ovarian tumors with good performance preoperatively. The mixed model with MLP classifier may be a potential tool in clinical applications.

Lane detection is a challenging task that requires predicting complex topology shapes of lane lines and distinguishing different types of lanes simultaneously. Earlier works follow a top-down roadmap to regress predefined anchors into various shapes of lane lines, which lacks enough flexibility to fit complex shapes of lanes due to the fixed anchor shapes. Lately, some works propose to formulate lane detection as a keypoint estimation problem to describe the shapes of lane lines more flexibly and gradually group adjacent keypoints belonging to the same lane line in a point-by-point manner, which is inefficient and time-consuming during postprocessing. In this paper, we propose a Global Association Network (GANet) to formulate the lane detection problem from a new perspective, where each keypoint is directly regressed to the starting point of the lane line instead of point-by-point extension. Concretely, the association of keypoints to their belonged lane line is conducted by predicting their offsets to the corresponding starting points of lanes globally without dependence on each other, which could be done in parallel to greatly improve efficiency. In addition, we further propose a Lane-aware Feature Aggregator (LFA), which adaptively captures the local correlations between adjacent keypoints to supplement local information to the global association. Extensive experiments on two popular lane detection benchmarks show that our method outperforms previous methods with F1 score of 79.63% on CULane and 97.71% on Tusimple dataset with high FPS. The code will be released at https://github.com/Wolfwjs/GANet., Accepted by CVPR2022

Geometry Projection is a powerful depth estimation method in monocular 3D object detection. It estimates depth dependent on heights, which introduces mathematical priors into the deep model. But projection process also introduces the error amplification problem, in which the error of the estimated height will be amplified and reflected greatly at the output depth. This property leads to uncontrollable depth inferences and also damages the training efficiency. In this paper, we propose a Geometry Uncertainty Projection Network (GUP Net) to tackle the error amplification problem at both inference and training stages. Specifically, a GUP module is proposed to obtains the geometry-guided uncertainty of the inferred depth, which not only provides high reliable confidence for each depth but also benefits depth learning. Furthermore, at the training stage, we propose a Hierarchical Task Learning strategy to reduce the instability caused by error amplification. This learning algorithm monitors the learning situation of each task by a proposed indicator and adaptively assigns the proper loss weights for different tasks according to their pre-tasks situation. Based on that, each task starts learning only when its pre-tasks are learned well, which can significantly improve the stability and efficiency of the training process. Extensive experiments demonstrate the effectiveness of the proposed method. The overall model can infer more reliable object depth than existing methods and outperforms the state-of-the-art image-based monocular 3D detectors by 3.74% and 4.7% AP40 of the car and pedestrian categories on the KITTI benchmark., Comment: To appear at ICCV2021

Additional file 1: table: Association between PVT1 expression and the subtypes of breast cancer patients (all female). Figure: si-PVT1 and sh-UPF1 knockdown efficiency in Hs578t and MCF-7 cells. (A) Western blot assays for FOXQ1 and UPF1 in Hs578t and MCF-7 cells. (B) Western blot assays for UPF1 in Hs578t and MCF-7 cells.

Developing new cabbage varieties resistant to head splitting and cabbage fusarium wilt (CFW) has become a high priority in breeding programs. To rapidly obtain resistant inbred lines, 230 doubled haploid (DH) lines were generated with an isolated microspore culture system from the crosses of two inbred lines, 23–457 (highly susceptible to both head splitting and CFW) and 88–54 (highly resistant to both head splitting and CFW). On the basis of field trials and a molecular marker assay, four elite DH lines (D29, D70, D120, and D162) were selected. These elite DH lines were crossed with elite cabbage inbred lines to generate more than 60 hybrids. An evaluation of the hybrids in the field indicated that five hybrids were highly resistant to both head splitting and CFW. These five hybrids also exhibited good agronomic traits. The resistant hybrids may be useful for controlling head splitting and CFW in the field. Our results also suggest that the application of an isolated microspore culture system and marker-assisted selection represents a rapid and efficient way of enhancing the resistance to head splitting and CFW in cabbage.

Epidermal growth factor (EGF)/EFG receptor (EGFR) signaling plays an important role in the osteoblastogenesis. The potential effects of pilose antler peptide (PAP) on osteoblast cell damages was investigated in our present study through EGF/EGFR signaling. In MC3T3-E1 osteoblastic cells, PAP treatment significantly inhibited the production of inflammatory cytokines by decreasing the levels of serum proinflammatory cytokines interleukin-1β (IL-1β), interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α). PAP treatment also alleviated the oxidative responses as indicated by increased activities of catalase (SOD) and decreased levels of malondialdehyde (MDA). EGF inhibition, by siRNA knockdown, almost abolished PAP-induced osteoblast cytoprotection against inflammation and oxidant stress. Further, our results showed that PAP stimulated the nuclear erythroid factor 2-related factor 2 (Nrf2)2/heme oxygenase-1(HO-1) signaling, and inhibited the activation of uclear factor kappa B (NF-κB) pathway in MC3T3-E1 cells. On the other hand, EGF siRNA knockdown inhibited PAP-induced cytoprotection, which decreased the expression of Nrf-2, HO-1 and increased the level of p-NF-κBp65, p-IκBα in MC3T3-E1 cells. Thus, our research demonstrated that PAP protects osteoblasts from inflammatory and oxidative injury through EGF/EGFR signaling.

In this study, we investigated anti-inflammatory effects of esculin (ESC) on lipopolysaccharide (LPS)-induced acute lung injury (ALI). ALI was induced in mice by intratracheal instillation of LPS, and ESC (20 and 40 mg/kg) was given orally 1 h prior to LPS administration. After 6 h, bronchoalveolar lavage fluid (BALF) and lung tissue were collected. ESC pretreatment decreased LPS-induced evident lung histopathological changes, lung wet-to-dry weight ratio, and lung myeloperoxidase activity. In addition, pretreatment with ESC inhibited inflammatory cells and proinflammatory cytokines including tumor necrosis factor-α (TNF-α), interleukin-1β, and interleukin-6 in BALF. Furthermore, we demonstrated that ESC inhibited the Toll-like receptor-2 (TLR2), Toll-like receptor-4 (TLR4), myeloid differentiation primary response gene-88 (MyD88), and nuclear factor-κB (NF-κB) p65 in LPS-induced ALI. The results indicated that the ESC had a protective effect on LPS-induced ALI in mice.

The aim of the current study was to use a mouse model of allergic asthma to investigate whether cordycepin has antiasthmatic effects, and if so, to determine the mechanism of these effects. A total of 50 mice were randomly assigned to five experimental groups: control, model, dexamethasone (Dex, 2 mg/kg), and cordycepin (20-40 mg/kg). Histological studies were evaluated by the hematoxylin and eosin staining, OVA-specific serum and BALF IgE levels and Treg/Th17 cytokines were evaluated by enzyme-linked immunosorbent assay, and RORγt and Foxp3 were evaluated by western blot. Our study demonstrated that cordycepin inhibited OVA-induced increases in eosinophil count; IL-17A levels were recovered and increased IL-10 levels in bronchoalveolar lavage fluid. Histological studies demonstrated that cordycepin substantially inhibited OVA-induced eosinophilia in lung tissue. Western blot study demonstrated that cordycepin increased Foxp3 and inhibited RORγt. These findings suggest that cordycepin may effectively ameliorate the progression of asthma and could be used as a therapy for patients with allergic asthma.

The present study was to evaluate the effect of ergosterol (ER) on CS (cigarette smoke)-induced chronic obstructive pulmonary disease (COPD) in mice. Fifty male ICR mice were randomly assigned to five groups: control group, CS group, CS + dexamethasone (Dex, 2 mg/kg) group, CS + ER (ER, 25 mg/kg) group, CS + ER (ER, 50 mg/kg). H&E staining demonstrated that ER inhibited CS-induced pathological injury in lung tissue. Besides, ER could restore the activities of superoxide dismutase (SOD) in serum and in the lung, catalase (CAT) in serum and reduce the content of malondialdehyde (MDA) in serum and in the lung. ER also inhibited pro-inflammatory cytokines including tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and interleukin-1β (IL-1β) in serum and the lung. Furthermore, ER significantly inhibited the protein expression of JAK3/STAT3/NF-κB pathway in CS-induced mice. Our findings suggested that ER might effectively ameliorate the progression of COPD via JAK3/STAT3/NF-κB pathway in mice.

In previous study, the anti-inflammatory effect of morin had been found. In this study, we investigated anti-inflammatory effects of morin on acute lung injury using lipopolysaccharide (LPS)-induced acute lung injury (ALI) mouse model. The cell counting in the bronchoalveolar lavage fluid (BALF) was measured. The animal lung edema degree was evaluated by wet/dry weight (W/D) ratio. The superoxidase dismutase (SOD) activity and myeloperoxidase (MPO) activity were assayed by SOD and MPO kits, respectively. The levels of inflammatory mediators including tumor necrosis factor-α (TNF-α), interleukin (IL)-1β, IL-18, and IL-6 were assayed by enzyme-linked immunosorbent assay method. Pathological changes of lung tissues were observed by hematoxylin and eosin (HE) staining. The protein level of lung NACHT, LRR, and PYD domains-containing protein 3 (NLRP3) inflammasome was measured by Western blotting. The data showed that treatment with the morin markedly attenuated inflammatory cell numbers in the BALF, decreased lung NLRP3 inflammasome protein level, and improved SOD activity and inhibited MPO activity. Histological studies demonstrated that morin substantially inhibited LPS-induced neutrophils in lung tissue compared with model group. The results indicated that the morin had a protective effect on LPS-induced ALI in mice.

Land consolidation engineering inevitably interferes with terrestrial ecosystems, leading to natural capital loss. Therefore, conducting an ecological sensitivity evaluation of a project area before consolidation engineering is very important for reducing unnecessary human interference. Conservation of terrestrial ecosystems and the biodiversity therein to the greatest possible extent is urgently needed. This research analyzes the interference by human activities caused by land consolidation engineering in terrestrial ecosystems. GIS technology, ecological values, landscape pattern indexes, and an ecological risk evaluation were used to construct an ecological sensitivity evaluation index. The coefficient of variation method and a comprehensive sensitivity rating evaluation were used to calculate the weights and results. The project area was divided into sensitivity zones according to the results, and the results and suggestions are as follows: In the highly eco-sensitive zone, where bare rocks, gravel, and grass-covered areas compose the main landscape type, vegetation should be restored, and forests should be planted. In the medium eco-sensitivity zone, where irrigated paddy fields and arid land compose the main landscape type, land parcels should be merged, and agricultural infrastructure should be constructed or improved. In the low eco-sensitivity zone, where forests compose the main landscape type, roads should be closed, natural habitat should be restored, and buffer zones should be created. This study provides suggestions for future land management decisions.

We consider a distributed Software Defined Networking (SDN) architecture adopting a cluster of multiple controllers to improve network performance and reliability. Besides the Openflow control traffic exchanged between controllers and switches, we focus on the control traffic exchanged among the controllers in the cluster, needed to run coordination and consensus algorithms to keep the controllers synchronized. We estimate the effect of the inter-controller communications on the reaction time perceived by the switches depending on the data-ownership model adopted in the cluster. The model is accurately validated in an operational Software Defined WAN (SDWAN). We advocate a careful placement of the controllers, that should take into account both the above kinds of control traffic. We evaluate, for some real ISP network topologies, the delay tradeoffs for the controllers placement problem and we propose a novel evolutionary algorithm to find the corresponding Pareto frontier. Our work provides novel quantitative tools to optimize the planning and the design of the network supporting the control plane of SDN networks, especially when the network is very large and in-band control plane is adopted. We also show that for operational distributed controllers (e.g. OpenDaylight and ONOS), the location of the controller which acts as a leader in the consensus algorithm has a strong impact on the reactivity perceived by switches., Comment: 14 pages

The extensive management mode of coal enterprises is no longer applicable to the demand of enterprise development under the new economic situation. Combined with the characteristics of coal mine production, based on the system of thinking, integration of lean, people, comprehensive, job management theory, formed HOPE management model, including a core system and three support systems and 18 elements. There are three stages in the development and implementation of this model. To 6S site management for the initial stage to job process reengineering for the intermediate stage to post value process control for the advanced stage. The successful implementation of HOPE model in coal enterprises needs comprehensive control from five aspects: lean culture construction, flattening organizational structure, cost control system, performance appraisal system and lean information management platform. HOPE model can be implemented smoothly and make "win-win" between enterprises and employees.