Complementary Branches Research Articles

Consistency label-activated region generating network for weakly supervised medical image segmentation

The current methods of auto-segmenting medical images are limited due to insufficient and ambiguous pathonmorphological labeling. In clinical practice, rough classification labels (such as disease or normal) are more commonly used than precise segmentation masks. However, there is still much to be explored regarding utilizing these weak clinical labels to accurately determine the lesion mask and guide medical image segmentation. In this paper, we proposed a weakly supervised medical image segmentation model to directly generate the lesion mask through a class activation map (CAM) guided cycle-consistency label-activated region transferring network. Cycle-consistency enforces that the mappings between the two domains should be reversible, which ensures that the original image can be reconstructed from the translated image. We developed a complementary branches fusion module to address the issue of blurry boundaries in CAM-guided segmentation. The complementary branch preserves the original semantic information of the non-lesion region and perfectly fuses the transferred feature of the lesion region with a complementary mask-constrained fake image generation process to clear the boundary of the lesion and non-lesion regions. This module allows the class transformation to focus solely on the label-activated region, resulting in more explicit segmentation. This model can accurately identify different region of medical images at the pixel-level while preserving the overall semantic structure semantion. It organizes disease labels and corresponding regions during image synthesis. Our method utilizes a joint discrimination strategy that significantly enhances the precision of the produced lesion mask. Extensive experiments of the proposed method on BraTs, ISIC and COVID-19 datasets demonstrate superior performance over existing state-of-the-art methods. The code and datasets are available at: https://github.com/mlcb-jlu/MedImgSeg.

Comparative analysis of CFD and ANFIS for predicting heat transfer enhancement in water-Fe2O3 nanofluids across various flow regions

Models for enhancement of heat transfer in nanofluids made wide use of adaptive neural fuzzy inference system (ANFIS) and the multi-phase mixture model in recent years. These models originate from two separate but complementary branches of engineering: computational mechanics and machine intelligence. Not only have prior studies used only a small subset of nanofluid and flow parameters in their analyses, but no one has ever compared the two methods to determine which one is more applicable to certain flow regimes to forecast how much heat transfer development nanofluids will exhibit. The purpose of this study was to compare the accuracy of two methods ? CFD and ANFIS in predicting the heat transfer improvement of water-Fe2O3 nanofluid for variety of nanofluid formations and flow characteristics, and recommend the method that would be most useful in predicting this enhancement for each flow regime. While ANFIS consistently outperforms the mixture models in prediction of nanofluid heat transfer enhancement, the latter can sometimes produce results that differ greatly from experimental correlation; however, for nanofluid configurations, the mixture model?s predictions can be dependable (with 1% error).

MuLHiTA: A Novel Multiclass Classification Framework With Multibranch LSTM and Hierarchical Temporal Attention for Early Detection of Mental Stress.

Mental stress is an increasingly common psychological issue leading to diseases such as depression, addiction, and heart attack. In this study, an early detection framework based on electroencephalogram (EEG) data is developed for reducing the risk of these diseases. In existing frameworks, signals are often segmented into smaller sections prior to being input to a deep neural network. However, this approach ignores the fundamental nature of EEG signals as a carrier of valuable information (e.g., the integrity of frequency and phase, and temporal fluctuations of EEG components). As such, this type of segmenting may lead to information loss and a failure to effectively identify mental stress levels. Thus, we propose a novel multiclass classification framework termed multibranch LSTM and hierarchical temporal attention (MuLHiTA) for the early identification of mental stress levels. It specifically focuses on not only intraslice (within each slice) but also interslice (between different slices) samples in parallel. This was achieved by including two complementary branches, each of which integrated a specifically designed attention module into a bidirectional long short-term memory (BLSTM) network, enabling extraction of the most discriminative features from interslice and intraslice EEG signals simultaneously. The outputs of attention modules were then summed to obtain a feature representation that contributes to reduce overfitting and more effective multiclass classification. In addition, electrode positions were optimized using neural activity areas under high-stress conditions, thereby reducing computational costs by minimizing the number of critical electrodes. MuLHiTA was evaluated across one private [Montreal imaging stress task (MIST)] and two publicly available EEG datasets [EEG during mental arithmetic tasks (DMAT) and Simultaneous task EEG workload (STEW)]. These were divided into training and test sets using an 8:2 ratio, and the training data were further divided into training and validation sets using a fivefold cross-validation (CV) method, in which the model with the highest accuracy among the five was selected. The model was trained once more with the full training set, and the test data were then used to evaluate its performance. This approach achieved average classification accuracies of 93.58%, 91.80%, and 99.71% for the MIST, STEW, and DMAT datasets, respectively. Experimental results showed MuLHiTA was superior to state-of-the-art algorithms, including EEGNet, BLSTM, EEGLearn, convolutional neural network (CNN)-long short-term memory (LSTM), and convolutional recurrent attention model (CRAM), for multiclass classification. This demonstrates the viability of MuLHiTA for the early detection of mental stress.

A dystopian dialogue – Semantic and pragmatic perspectives on George Orwell’s 1984 and Boualem Sansal’s 2084

Dystopias are no longer defined as the opposite of utopias and are mostly described by referring to the type of system they depict - “perfectly planned and beneficial”, “perfectly planned and unjust” or “perfectly unplanned” (Gordin, Tilley, and Prakash 2010, 2). Furthermore, dystopias are said to reflect societies’ worries and warn about some flaws of religious, political systems, or science in terms of relationships between the past, the present, and the future. George Orwell’s “1984” and Boualem Sansal’s “2084”. The End of the World are dystopian novels that fall into two different categories (based on the type of authority exercising control) – the former is a political dystopia and the latter is a religious one. 2084 has often been described as a tribute to Orwell and his ‘Big Brother’, but the cult of personality is depicted by pointing at elements related to religion. It stands out as an even more explicit illustration of totalitarian regimes and their practices especially when connected to religion. However, both novels revolve around the concepts of ideology, the cult of personality, and power enforcement. Thus, despite their typology, the dialogue between these two dystopias becomes obvious as regards genre and intertextuality, but also the semantic and pragmatic features they share. Therefore, with a view to tracing the meanings conveyed through language, the present paper tackles the aforementioned dystopian novels from the perspective of these two complementary branches of linguistics, in an attempt at identifying the similarities (especially those referring to ideologies and power).

VUCA environments before the recession caused by covid-19: a systematic literature review

The acronym VUCA as a description of a Volatile, Uncertain, Complex and Ambiguous environment, has been gaining increasing relevance as a noun to describe a reality that is complex and turbulent. Besides the healthcare challenges, the COVID-19 pandemic has also caused drastic changes in the management world. Through a systematic literature review of 137 articles published before the beginning of the COVID-19 pandemic (between 2009 and March 2020), we develop a conceptual framework that integrates complementary branches of theoretical and empirical research, with the VUCA environment as the central unit of analysis. Our objectives are to unravel how the main theories in this field address essential aspects of management and future study trends, we also propose a conceptualisation of VUCA as an essential element of the current environment and its relationship with the different disciplines in the business world.

Rethinking Lightweight Salient Object Detection via Network Depth-Width Tradeoff.

Existing salient object detection methods often adopt deeper and wider networks for better performance, resulting in heavy computational burden and slow inference speed. This inspires us to rethink saliency detection to achieve a favorable balance between efficiency and accuracy. To this end, we design a lightweight framework while maintaining satisfying competitive accuracy. Specifically, we propose a novel trilateral decoder framework by decoupling the U-shape structure into three complementary branches, which are devised to confront the dilution of semantic context, loss of spatial structure and absence of boundary detail, respectively. Along with the fusion of three branches, the coarse segmentation results are gradually refined in structure details and boundary quality. Without adding additional learnable parameters, we further propose Scale-Adaptive Pooling Module to obtain multi-scale receptive field. In particular, on the premise of inheriting this framework, we rethink the relationship among accuracy, parameters and speed via network depth-width tradeoff. With these insightful considerations, we comprehensively design shallower and narrower models to explore the maximum potential of lightweight SOD. Our models are proposed for different application environments: 1) a tiny version CTD-S (1.7M, 125FPS) for resource constrained devices, 2) a fast version CTD-M (12.6M, 158FPS) for speed-demanding scenarios, 3) a standard version CTD-L (26.5M, 84FPS) for high-performance platforms. Extensive experiments validate the superiority of our method, which achieves better efficiency-accuracy balance across five benchmarks.

ПЕРСПЕКТИВЫ ПРИМЕНЕНИЯ ПОТЕНЦИАЛА НЕЙРОМАРКЕТИНГА

Neuromarketing is one of the new and modern types of marketing, which has a multifaceted specificity, a wide scope of application, has found its target audience and has gained its popularity as an effective technology for managing consumer behavior. It is believed that neuromarketing is the area of formation and action of physiological and neural signals that affect the motives, tastes and preferences of the target audience. The potential of neuromarketing includes a set of special procedures and methods that, with the help of specific operations, fix the necessary information in the minds of consumers and emotionally involve them in the process of promoting goods and services. In this way, the brain's reaction to advertising and other messages about a product or service is recorded, and on the basis of neuroexamination, consumer behavior and his reaction to a new product are evaluated and tracked. The features of neuromarketing and its potential are studied, the prospects of application are determined. The conclusion is made: the symbiosis of complementary branches of science with the effective implementation of the set goals will continue to play an important role in the system of promotion of goods and services.

Collaborative learning network for head pose estimation

• Propose a collaborative learning framework for head pose estimation. • Learn complementary information from the landmark-based and landmark-free branches. • Design a Landmark-MLP-Mixer to obtain the head pose angles from facial landmarks. • Introduce a union loss function to optimize two complementary branches. • Conduct comparison experiments and ablation studies on various datasets. Head pose estimation is an important task in many real-world applications, such as human–computer interaction, driver monitoring, face localization and gaze estimation. In this paper, we present a novel collaborative learning framework based on Convolutional Neural Networks (CNNs) for head pose estimation. The proposed framework consists of a landmark-based branch and a landmark-free branch. The former first estimates facial landmarks and then follows the Landmark-MLP-Mixer module which models the complex nonlinear mapping relationship from facial landmarks to head pose angles. While the later adopts a label distribution learning strategy to estimate head pose. The two branches both dedicate themselves to head pose estimation task, and they collaborate with each other for mutual promotion and complementary semantic learning. Specifically, we introduce a dual-branch transfer module in the middle of the network to achieve explicit semantic interaction and introduce a multi-loss strategy that induces to implicit information interaction. We conduct extensive experiments on several popular benchmarks, including AFLW, AFLW2000 and BIWI, the results show that our method is competitive compared to other state-of-the-art methods.

A Disentangled Representation Based Brain Image Fusion via Group Lasso Penalty.

Complementary and redundant relationships inherently exist between multi-modal medical images captured from the same brain. Fusion processes conducted on intermingled representations can cause information distortion and the loss of discriminative modality information. To fully exploit the interdependency between source images for better feature representation and improve the fusion accuracy, we present the multi-modal brain medical image fusion method in a disentangled pipeline under the deep learning framework. A three-branch auto-encoder with two complementary branches and a redundant branch is designed to extract the exclusive modality features and common structure features from input images. Especially, to promote the disentanglement of complement and redundancy, a complementary group lasso penalty is proposed to constrain the extracted feature maps. Then, based on the disentangled representations, different fusion strategies are adopted for complementary features and redundant features, respectively. The experiments demonstrate the superior performance of the proposed fusion method in terms of structure preservation, visual quality, and running efficiency.

RPformer: A Robust Parallel Transformer for Visual Tracking in Complex Scenes

The Siamese architecture has shown remarkable performance in the field of visual tracking. Although the existing Siamese-based tracking methods have achieved a relative balance between accuracy and speed, the performance of many trackers in complex scenes is often unsatisfactory, which is mainly caused by interference factors, such as target scale changes, occlusion, and fast movement. In these cases, excessive trackers cannot employ sufficiently the target feature information and face the dilemma of information loss. In this work, we propose a novel parallel Transformer network architecture to achieve robust visual tracking. The proposed method designs the Transformer-1 module, the Transformer-2 module, and the feature fusion head (FFH) based on the attention mechanism. The Transformer-1 module and the Transformer-2 module are regarded as corresponding complementary branches in the parallel architecture. The FFH is used to integrate the feature information of the two parallel branches, which can efficiently exploit the feature dependence relationship between the template and the search region, and comprehensively explore rich contextual information. Finally, by combining the core ideas of Siamese and Transformer, we present a simple and robust tracking framework called RPformer, which does not require any prior knowledge and avoids the trouble of adjusting hyperparameters. Numerous experiments show that the proposed tracking method achieves more outstanding performance than the state-of-the-art trackers on seven tracking benchmarks, which can meet the real-time requirements at a running speed exceeding 50.0 frames/s.

Spatial Dynamic Selection Network for Remote-Sensing Image Fusion

Nowadays, high-resolution images with rich spectral information are necessary for earth observation. Remote-sensing image fusion is an effective method to provide high-resolution multispectral images, which are obtained by fusing high-resolution panchromatic images and low-resolution multispectral images. However, existing methods mostly use the same network for image feature extraction, without considering the differences among different pixels, resulting in that the extracted features are not accurate enough. This letter proposes a spatial dynamic selection network for remote-sensing image fusion. A dynamic feature extraction module composed of multiple spatial dynamic blocks (SDBs) and cross-scale context connection blocks (CSCBs) is designed. The SDB can extract image features according to the input by different networks, and realize dynamic selection of pixel features. Since the spatial structure and spectral characteristic of each pixel are different, two complementary branches are designed in the SDB to extract different features, which improves the capability of feature extraction. Multiscale network structure is designed to obtain more abundant information and the CSCB is used to integrate the information of different scales. Experimental results on GeoEye-1 and WorldView-3 datasets demonstrate the superiority of the proposed method.

A Dual-Branch Self-Boosting Framework for Self-Supervised 3D Hand Pose Estimation.

Although 3D hand pose estimation has made significant progress in recent years with the development of the deep neural network, most learning-based methods require a large amount of labeled data that is time-consuming to collect. In this paper, we propose a dual-branch self-boosting framework for self-supervised 3D hand pose estimation from depth images. First, we adopt a simple yet effective image-to-image translation technology to generate realistic depth images from synthetic data for network pre-training. Second, we propose a dual-branch network to perform 3D hand model estimation and pixel-wise pose estimation in a decoupled way. Through a part-aware model-fitting loss, the network can be updated according to the fine-grained differences between the hand model and the unlabeled real image. Through an inter-branch loss, the two complementary branches can boost each other continuously during self-supervised learning. Furthermore, we adopt a refinement stage to better utilize the prior structure information in the estimated hand model for a more accurate and robust estimation. Our method outperforms previous self-supervised methods by a large margin without using paired multi-view images and achieves comparable results to strongly supervised methods. Besides, by adopting our regenerated pose annotations, the performance of the skeleton-based gesture recognition is significantly improved.

Learning event guided network for salient object detection

• We propose an end-to-end network for salient object detection by combining the asynchronous event data and the RGB image. • We utilize Long Short-Term Memory and multi-level feature interaction modules to complement the image and event data. • We build an Event-RGB SOD dataset (ERSOD), and make it available to the research community. Salient object detection (SOD) focuses on mimicking the attention mechanism in human vision system. Due to the limited information provided from the traditional image data, the image-based SOD advances are very challenging in some complex scenes. Inspired by the emerging event cameras that provide asynchronous measurements of local temporal contrast over a large dynamic range, we propose a new idea to extract more effective information from the combination of event flow and RGB images. In this paper, we construct an end-to-end joint network for salient object detection (ERSOD-Net), which simultaneously supervise the RGB image and the event data within the corresponding image exposure time. To fully exploit temporal information of event data, Long Short-Term Memory module is utilized to effectively process event and learn salient object event surfaces. Moreover, multi-level feature interaction is designed to fuse two complementary branches of image and event and to predict significance mapping. Finally, to demonstrate the effectiveness of our model, a real Event-RGB SOD dataset (ERSOD) is built by DAVIS camera. Experiments on both benchmark and ERSOD datasets show that the proposed event guided network greatly improves the SOD performance in different evaluation metrics. The code and datasets will be released later at https://github.com/jxr326/ERSOD-Net.

HF-UNet: Learning Hierarchically Inter-Task Relevance in Multi-Task U-Net for Accurate Prostate Segmentation in CT Images.

Accurate segmentation of the prostate is a key step in external beam radiation therapy treatments. In this paper, we tackle the challenging task of prostate segmentation in CT images by a two-stage network with 1) the first stage to fast localize, and 2) the second stage to accurately segment the prostate. To precisely segment the prostate in the second stage, we formulate prostate segmentation into a multi-task learning framework, which includes a main task to segment the prostate, and an auxiliary task to delineate the prostate boundary. Here, the second task is applied to provide additional guidance of unclear prostate boundary in CT images. Besides, the conventional multi-task deep networks typically share most of the parameters (i.e., feature representations) across all tasks, which may limit their data fitting ability, as the specificity of different tasks are inevitably ignored. By contrast, we solve them by a hierarchically-fused U-Net structure, namely HF-UNet. The HF-UNet has two complementary branches for two tasks, with the novel proposed attention-based task consistency learning block to communicate at each level between the two decoding branches. Therefore, HF-UNet endows the ability to learn hierarchically the shared representations for different tasks, and preserve the specificity of learned representations for different tasks simultaneously. We did extensive evaluations of the proposed method on a large planning CT image dataset and a benchmark prostate zonal dataset. The experimental results show HF-UNet outperforms the conventional multi-task network architectures and the state-of-the-art methods.

The Role of Slovenian Local Government in the Wine Tourism Promotion

Wine, viticulture, and winemaking are an important cornerstone of economic development, culture, and tourism. Especially in wine-growing parts of the world where symbols related to wine are used even as local or national symbols. Viticulture and its complementary branches are differently developed in different parts of the world – the predisposition of further development, however, is also influenced importantly by positions of different identification referential groups which they have towards the use of wine and further development of the profession. Based on the research of standpoints of the mayors in the Republic of Slovenia, which could be presented as “wine-growing country” with regards to the extent of the vine, we can assert that mayors with their relationship towards wine and their operations influence the use of local wine for the tourist promotion of destination importantly. According to that, we figure out that mayors play an important role in the development of “public policy” in the local environment. In the forming of the positions towards the use of wine for the promotion, the direct experiences are important, as well as the environment where the mayors come from and all identification elements in the sense of cultural heritage, tradition, and development of the branch.

A Complementary Recycling Operational Transconductance Amplifier with Data-Driven Enhancement of Transconductance

An improved operational transconductance amplifier (OTA) is presented in this work. The fully differential OTA adopts the current recycling technique and complementary NMOS and PMOS input branches to enhance the total transconductance. Moreover, in order to achieve higher current efficiency, a data-driven biasing circuit was developed to dynamically adjust the power consumption of the amplifier. Two comparators were added to detect the voltage difference at the input nodes, and when the differential input is large enough to activate either comparator, extra biasing current is activated and poured into the amplifier to enhance its slew rate and gain-bandwidth product (GBW). The threshold voltage of the complementary recycling folded cascode (CRFC)-based comparator is configured to suppress overshoot. Complementary common-mode feedback (CMFB) topology with local CMFB structure is built to acquire high common-mode gain. The OTA was fabricated in SMIC 0.18- μ m CMOS technology. The experimental result based on a capacitive feedback loop shows that the data-driven operation improves the average slew rate of the amplifier from 10.2 V/ μ s to 55.5 V/ μ s while the power only increases by 150%. The OTA has good potential to satisfy the fast settling demands for capacitive sensing circuits.

Modeling of a supercapacitor with a multibranch circuit

The modeling of a supercapacitor by an electrical circuit composed of multiple resistors and variable voltage capacitors is studied. The temporal characterization is performed with a realized test bench with accurate data acquisition for the different charge and discharge protocols. Access resistance and RC branches are identified for the different time windows. The experimental characterization is detailed for the transmission line and for the complementary branches which reproduce redistribution and leakage current. Thereafter, the implementation of the equivalent circuit in the Matlab/Simulink environment is shown, namely the consideration of non-linear behaviors of the capacitances regarding voltage. The found characterized values are entered in the model to simulate charge/discharge of the supercapacitor. Agreement between simulated and experimental results proves the consistency of the time constant determination approach.

Contributions of charge-density research to medicinal chemistry.

This article reviews efforts in accurate experimental charge-density studies with relevance to medicinal chemistry. Initially, classical charge-density studies that measure electron density distribution via least-squares refinement of aspherical-atom population parameters are summarized. Next, interaction density is discussed as an idealized situation resembling drug-receptor interactions. Scattering-factor databases play an increasing role in charge-density research, and they can be applied both to small-molecule and macromolecular structures in refinement and analysis; software development facilitates their use. Therefore combining both of these complementary branches of X-ray crystallography is recommended, and examples are given where such a combination already proved useful. On the side of the experiment, new pixel detectors are allowing rapid measurements, thereby enabling both high-throughput small-molecule studies and macromolecular structure determination to higher resolutions. Currently, the most ambitious studies compute intermolecular interaction energies of drug-receptor complexes, and it is recommended that future studies benefit from recent method developments. Selected new developments in theoretical charge-density studies are discussed with emphasis on its symbiotic relation to crystallography.

Position and sequence conservation in Amniota of polymorphic enhancer HS1.2 within the palindrome of IgH 3'Regulatory Region

BackgroundThe Immunoglobulin heavy chain (IgH) 3' Regulatory Region (3'RR), located at the 3' of the constant alpha gene, plays a crucial role in immunoglobulin production. In humans, there are 2 copies of the 3'RR, each composed of 4 main elements: 3 enhancers and a 20 bp tandem repeat. The single mouse 3'RR differs from the two human ones for the presence of 4 more regulative elements with the double copy of one enhancer at the border of a palindromic region.ResultsWe compared the 3'RR organization in genomes of vertebrates to depict the evolutionary history of the region and highlight its shared features. We found that in the 8 species in which the whole region was included in a fully assembled contig (mouse, rat, dog, rabbit, panda, orangutan, chimpanzee, and human), the shared elements showed synteny and a highly conserved sequence, thus suggesting a strong evolutionary constraint. In these species, the wide 3'RR (~30 kb in human) bears a large palindromic sequence, consisting in two ~3 kb complementary branches spaced by a ~3 kb sequence always including the HS1.2 enhancer. In mouse and rat, HS3 is involved by the palindrome so that one copy of the enhancer is present on each side. A second relevant feature of our present work concerns human polymorphism of the HS1.2 enhancer, associated to immune diseases in our species. We detected a similar polymorphism in all the studied Catarrhini (a primate parvorder). The polymorphism consists of multiple copies of a 40 bp element up to 12 in chimpanzees, 8 in baboons, 6 in macaque, 5 in gibbons, 4 in humans and orangutan, separated by stretches of Cytosine. We show specific binding of this element to nuclear factors.ConclusionsThe nucleotide sequence of the palindrome is not conserved among evolutionary distant species, suggesting pressures for the maintenance of two self-matching regions driving a three-dimensional structure despite of the inter-specific divergence at sequence level. The information about the conservation of the palindromic structure and the settling in primates of the polymorphic feature of HS1.2 show the relevance of these structures in the control and modulation of the Ig production through the formation of possible three-dimensional structures.

Mathematics and Artificial Intelligence, two branches of the same tree

Unfortunately, in the learning of Mathematics and Computer Science, they appear often as disconnected areas, when they are indeed two necessary and complementary branches of the same tree. Either of them alone produces only ethereal structures, or routines and ad-hoc programs. For this reason, it would be preferable to study, progressively, from the lower educational levels, both disciplines as naturally linked. So, it will be overrated the pure mechanistic of only give informatics to usuary level, as mere blind instructions, either too abstract pure mathematical constructs.