A new lightweight network for efficient UAV object detection.

Optimizing the structure of deep neural networks is essential in many applications. Especially in the object detection tasks of Unmanned Aerial Vehicles. Due to the constraints of the onboard platform, a more efficient network is required to meet practical demands. Nevertheless, existing lightweight detection networks exhibit excessive redundant computations and may yield in a certain level of accuracy loss. To address these issues, this paper proposes a new lightweight network structure named Cross-Stage Partially Deformable Network (CSPDNet). The initial proposal consists of a Deformable Separable Convolution Block (DSCBlock), separating feature channels, greatly reducing the computational load of convolution, and applying adaptive sampling to the separated feature map. Subsequently, to establish information interaction between feature layers, a channel weighting module is proposed. This module calculates weights for the separated feature map, facilitating information exchange across channels and resolutions. Moreover, it compensates for the effect of point-wise (1 × 1) convolutions, filtering out more important feature information. Furthermore, a new CSPDBlock is designed, primarily composed of DSCBlock, establishing multidimensional feature correlations for each separated feature layer. This approach improves the ability to capture critical feature information and reconstruct gradient paths, thereby preserving detection accuracy. The proposed technology achieves a balance between model parameter size and detection accuracy. Furthermore, experimental results on object detection datasets demonstrate that our designed network, using fewer parameters, achieves competitive detection performance results compared to existing lightweight networks YOLOv5n, YOLOv6n, YOLOv8n, NanoDet and PP-PicoDet. The optimization effect of the designed CSPDBlock, using the VisDrone dataset, is validated when incorporated into advanced detection algorithms YOLOv5m, PPYOLOEm, YOLOv7, RTMDetm and YOLOv8m. In more detail, by incorporating the designed modules it was achieved that the parameters were reduced by 10-20% while almost maintaining detection accuracy.

[Congenital Fâ ¦ Deficiency Associated with a Novel Mutation in F7 Gene].

OBJECTIVE: To identify the genetic mutation of coagulation factor Ⅶ ( F7) gene in a pedigree with coagulation factor Ⅶ (FⅦ) deficiency and explore the molecular pathogenesis. METHODS: The prothrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), D-dimer (DD), fibrin degradation products (FDP) and coagulation factor Ⅶ activity (FⅦ:C) of the proband and her family members were detected by Sysmex-CS5100 analyzer. All exons and exon-intron boundaries of the F7 gene were amplified by PCR followed by direct sequencing. The detected mutation was confirmed by reverse sequencing. The ClustalW software was used to analyze the conservatism of the mutant site. Pathogenicity of the mutation was assessed with Mutation Taster and PolyPhen-2 online bioinformatics software. Structure of the mutant protein was analyzed using Swiss-PdbViewer software. RESULTS: The results of routine coagulation tests showed that PT of the proband was markedly extended to 42.5 s, and her FⅦ:C significantly reduced to 2%. The FⅦ:C of her grandmother, mother and sister had slightly reduced to 49%, 51%, and 42%, respectively. These coagulant parameters of her father were within the normal range. Genetic analysis reveled a heterozygous G>A change at cDNA 646 in exon 6 of F7 gene in the proband, resulting in a replacement of glycine at 156 of FⅦ catalytic region with serine (p.Gly156Ser). The sequencing results of other exons and exon-intron boundaries of her F7 gene were normal. The proband's grandmother, mother and sister were all the carriers of this missense mutation except her father. Bioinformatics analysis showed that the p.Gly156Ser mutation caused polarity change of the amino acid at this site and formation of side chains, leading to increase of protein instability, which may affect catalytic activity of structural domain. Meanwhile, both Mutation Taster and PolyPhen-2 online bioinformatics software also predicted the pathogenicity of this missense mutation with high scores. CONCLUSION: The heterozygous p.Gly156Ser mutation is the direct cause of the reduced FⅦ in this proband.

Semi-Supervised Detection Model Based on Adaptive Ensemble Learning for Medical Images.

Introducing deep learning technologies into the medical image processing field requires accuracy guarantee, especially for high-resolution images relayed through endoscopes. Moreover, works relying on supervised learning are powerless in the case of inadequate labeled samples. Therefore, for end-to-end medical image detection with overcritical efficiency and accuracy in endoscope detection, an ensemble-learning-based model with a semi-supervised mechanism is developed in this work. To gain a more accurate result through multiple detection models, we propose a new ensemble mechanism, termed alternative adaptive boosting method (Al-Adaboost), combining the decision-making of two hierarchical models. Specifically, the proposal consists of two modules. One is a local region proposal model with attentive temporal-spatial pathways for bounding box regression and classification, and the other one is a recurrent attention model (RAM) to provide more precise inferences for further classification according to the regression result. The proposal Al-Adaboost will adjust the weights of labeled samples and the two classifiers adaptively, and the nonlabel samples are assigned pseudolabels by our model. We investigate the performance of Al-Adaboost on both the colonoscopy and laryngoscopy data coming from CVC-ClinicDB and the affiliated hospital of Kaohsiung Medical University. The experimental results prove the feasibility and superiority of our model.

A Dimensionality-Reducible Operational Optimal Control for Wastewater Treatment Process.

Operational optimal control (OOC) is an essential component of wastewater treatment process (WWTP). The control variables usually are high-dimensional, nonlinear, and strongly coupled, which can easily fail traditional optimization control methods. Mathematically, these operational variables usually are in the unknown low-dimensional space embedded in the high-dimensional space. Therefore, the OOC problem of WWTP can be resolved as an optimization challenge involving low-dimensional space, and the unknown low-dimensional space is presented in the form of a set of controlled variables in a high-dimensional space, which is normal in real-world industries. Here, a dimension-reducible data-driven optimization control framework for WWTP is proposed. Considering the difficulty in elucidating the whole space of set points, a neural network is designed to approximate the constraint relationship between control variables. The search process is based on optimization methods in low-dimensional space embedded into Euclidean spaces. Furthermore, the convergence of the process is ensured via mathematical analysis. Finally, the experimental simulation of wastewater treatment revealed that this approach is effective for an optimal solution in control systems.

Two-Stream Retentive Long Short-Term Memory Network for Dense Action Anticipation.

Analyzing and understanding human actions in long-range videos has promising applications, such as video surveillance, automatic driving, and efficient human-computer interaction. Most researches focus on short-range videos that predict a single action in an ongoing video or forecast an action several seconds earlier before it occurs. In this work, a novel method is proposed to forecast a series of actions and their durations after observing a partial video. This method extracts features from both frame sequences and label sequences. A retentive memory module is introduced to richly extract features at salient time steps and pivotal channels. Extensive experiments are conducted on the Breakfast data set and 50 Salads data set. Compared to the state-of-the-art methods, the method achieves comparable performance in most cases.

Discontinuous Track Recognition System Based on PolyLaneNet for Darwin-op2 Robot.

This paper proposes and demonstrates a single-line discontinuous track recognition system by associating the track recognition problem of a humanoid robot with the lane detection problem. The proposal enables the robot to achieve stable running on the single-line discontinuous track. The system consists of two parts: the robot end and the graphics computing end. The robot end is responsible for collecting track information and the graphics computing end is responsible for high-performance computing. These two parts use the TCP for communication. The graphics computing side uses PolyLaneNet lane detection algorithm to train the track image captured from the first perspective of the darwin-op2 robot as the data set. In the inference, the robot end sends the collected tracking images to the graphics calculation end and uses the graphics processor to accelerate the calculation. After obtaining the motion vector, it is transmitted back to the robot end. The robot end parses the motion vector to obtain the motion information of the robot so that the robot can achieve stable running on the single-line discontinuous track. The proposed system realizes the direct recognition of the first perspective image of the robot and avoids the problems of poor stability, inability of identifying curves and discontinuous lines, and other problems in the traditional line detection method. At the same time, this system adopts the method of cooperative work between the PC side and the robot by deploying the algorithm with high computational requirements on the PC side. The data transmission is carried out by stable TCP communication, which makes it possible for the robot equipped with weak computational controllers to use deep-learning-related algorithms. It also provides ideas and solutions for deploying deep-learning-related algorithms on similar low computational robots.

Graph-Based Visual Manipulation Relationship Reasoning Network for Robotic Grasping.

To grasp the target object stably and orderly in the object-stacking scenes, it is important for the robot to reason the relationships between objects and obtain intelligent manipulation order for more advanced interaction between the robot and the environment. This paper proposes a novel graph-based visual manipulation relationship reasoning network (GVMRN) that directly outputs object relationships and manipulation order. The GVMRN model first extracts features and detects objects from RGB images, and then adopts graph convolutional network (GCN) to collect contextual information between objects. To improve the efficiency of relation reasoning, a relationship filtering network is built to reduce object pairs before reasoning. The experiments on the Visual Manipulation Relationship Dataset (VMRD) show that our model significantly outperforms previous methods on reasoning object relationships in object-stacking scenes. The GVMRN model is also tested on the images we collected and applied on the robot grasping platform. The results demonstrated the generalization and applicability of our method in real environment.

Exosomal ephrinA2 derived from serum as a potential biomarker for prostate cancer.

Up-regulation of serum ephrinA2 is common in various malignancies and has been suggested as a potential biomarker for the diagnosis and prognosis of prostate cancer (PCa). However, the type of serum ephrinA2 expressed in PCa patients remains elusive. Furthermore, the level of exosomal ephrinA2 derived from serum is increased in patients with osteoporosis, a common complication of PCa patients undergoing androgen deprivation therapy. It is unknown whether exosomes derived from PCa patient serum contains ephrinA2. In this study, we explored the ephrinA2 expression in whole serum and tissues and identified the circulating exosomal ephrinA2 as a potential biomarker for PCa. Exosomes were isolated from patient sera by differential centrifugation and the presence of ephrinA2 was confirmed via electron microscopy and western blotting. The type of ephrinA2 in serum was evaluated by western blotting. The expression of serum ephrinA2 including secreted and cleaved ephrinA2 and exosomal ephrinA2 were detected by ELISA and western blotting. Compared with benign prostatic hyperplasia (BPH) and controls, the levels of whole serum ephrinA2 and exosomal ephrinA2 were significantly higher in PCa patients. Moreover, exosomal ephrinA2 expression was positively correlated with TNM staging and Gleason score of PCa patients. The diagnostic efficiency of exosomal ephrinA2 was superior to that of whole serum ephrinA2 and serum PSA in distinguishing PCa patients from those from BPH patents. Our study indicates that exosomal ephrinA2 has high potential as a biomarker for the presence of PCa and offers a new therapeutic target for this disease.

Clinical and molecular characterization of nine Chinese patients affected by hypofibrinogenemia or dysfibrinogenemia.

: Congential fibrinogen deficiency is a rare bleeding disorder caused by various mutations in three fibrinogen genes. It can be subdivided into four categories: afibrinogenemia, hypofibrinogenemia, hypodysfibrinogenemia and dysfbrinogenemia. This study was to elucidate the molecular defects in nine unrelated Chinese patients with hypofibrinogenemia or dysfibrinogenemia. Three fibrinogen genes were amplified by PCR and screened for variants. The identified variants were analyzed by bioinformatics prediction and molecular modeling analysis. Genetic screening disclosed seven different missense mutations, four of which were novel. All of the mutations were expected to impair the protein function/structure as assessed by bioinformatics prediction. This study has increased our knowledge of the mutational spectrum underlying fibrinogen deficiency.