Recovery Block Research Articles

Multimodal Score Fusion with Sparse Low-rank Bilinear Pooling for Egocentric Hand Action Recognition

With the advent of egocentric cameras, there are new challenges where traditional computer vision is not sufficient to handle this kind of video. Moreover, egocentric cameras often offer multiple modalities that need to be modeled jointly to exploit complimentary information. In this article, we propose a sparse low-rank bilinear score pooling approach for egocentric hand action recognition from RGB-D videos. It consists of five blocks: a baseline CNN to encode RGB and depth information for producing classification probabilities; a novel bilinear score pooling block to generate a score matrix; a sparse low-rank matrix recovery block to reduce redundant features, which is common in bilinear pooling; a one-layer CNN for frame-level classification; and an RNN for video-level classification. We proposed to fuse classification probabilities instead of traditional CNN features from RGB and depth modality, involving an effective yet simple sparse low-rank bilinear score pooling to produce a fused RGB-D score matrix. To demonstrate the efficacy of our method, we perform extensive experiments over two large-scale hand action datasets, namely, THU-READ and FPHA, and two smaller datasets, GUN-71 and HAD. We observe that the proposed method outperforms state-of-the-art methods and achieves accuracies of 78.55% and 96.87% over the THU-READ dataset in cross-subject and cross-group settings, respectively. Further, we achieved accuracies of 91.59% and 43.87% over the FPHA and Gun-71 datasets, respectively.

An Efficient Image Compression Technique using Long Short-Term Memory Networks (LSTM)

The emergence of big data has imposed significant challenges on data storage and transmission. One pressing issue is leveraging deep learning techniques to achieve superior compression ratios and enhance image quality. Recurrent Neural Networks (RNNs) offer a promising avenue for controlling image bit rates iteratively, thereby enhancing compression performance. However, integrating Long Short-Term Memory (LSTM) into RNNs to address long-term dependencies increases model complexity. To expedite training and enhance image reconstruction quality, this study proposes several innovations.Initially, we enhance the activation function within LSTM to more effectively manage information retention and omission, thereby reducing parameter count and expediting training. Additionally, we introduce an image recovery block within the decoder to reconstruct high-resolution images. Finally, to expedite loss convergence, we replace L1 loss with SmoothL1 loss. Experimental outcomes demonstrate the efficacy of our approach, showcasing higher compression ratios

CURE-Net: A Cascaded Deep Network for Underwater Image Enhancement

Underwater imaging is usually degraded by low contrast or color cast due to light absorption and scattering. This fact limits the accuracy of object recognition in underwater and marine environments. In this article, to address this issue, we propose a CURE-Net that progressively improves degraded underwater images in a coarse-to-fine way. To be specific, our CURE-Net consists of three cascaded subnetworks. The first two subnetworks use attention and gate fusion to learn multiscale contextual information, whereas the third subnetwork preserves fine spatial details. A detail enhancement block and a supervised recovery block are deployed between subnetworks to further improve the recovery of colors and details. Our proposed CURE-Net is extensively evaluated on popular data sets and achieves state-of-the-art performance. Compared with its peers, our method is also more friendly to object recognition—underwater objects are more easily recognized after being processed by our method.

Thermal plasma-driven looping for metal scrap processing with hydrogen

In the present research, a novel process was proposed and assessed using an equilibrium thermochemical modelling study to evaluate the performance of a combined disruptive technology of thermal plasma and thermochemical looping to utilise hydrogen for iron reduction processing. The study was aimed at conducting a high-level power performance assessment of the process by integrating a built-in power block with a focus on a novel equilibrium thermal plasma technology for reducing iron scrap particles. The effect of thermal plasma operating parameters such as temperature (1000 °C < T < 1800 °C) and hydrogen-to-iron ratio in the feed stream to the thermal plasma reactor (0 < H2/metal < 20) on the performance of the system was numerically investigated using Aspen coupled with Matlab. Also, the self-sustaining factor was evaluated coupled with a quantitative analysis of the sustainability and lifecycle of the process aiming at better understanding the impact of the proposed system on the environment. Results showed that the system is crucially sensitive to the ratio of H2/metal such that the overall power demand of the reactor can change in a way that the dominant regime of the plasma can change from endothermic (partial reduction of iron) to exothermic (complete reduction, combustion of iron) once the hydrogen to metal ratio exceeds ∼ 2.5. It was also identified that an increase in the temperature of the combustor decreased the power demand for the thermal plasma reactor. Similarly, with increasing the temperature, a ramp-up in the self-sustaining factor was observed reaching 0.8 showing that 80 % of the energy of the auxiliaries can be maintained using the built-in heat recovery and power block. The thermal efficiency of the system was also a strong function of the H2/metal ratio reaching ∼ 0.4 at the H2/metal ratio of ∼ 2.5 reflecting the fact that the proposed process efficiency is within the state-of-the-art power production systems. The sustainability assessments showed that the process offers a high circular economy capability of the process with above-the-average lifetime and Material Circularity Indicator values.

10276 Quadratus Lumborum Block and Quality of Recovery Following Minimally Invasive Hysterectomy: A Randomized Double-Blind, Controlled Trial

10276 Quadratus Lumborum Block and Quality of Recovery Following Minimally Invasive Hysterectomy: A Randomized Double-Blind, Controlled Trial

Priors Guided Extreme Underwater Image Compression for Machine Vision and Human Vision

A low bit-rate compression is required for underwater images due to the limited bandwidth in underwater acoustic communication, which limits performances of both machine analysis and human perception in most underwater applications. Few existing compression methods consider the unique characteristics of underwater images such as color shift and haze effect to better fulfill requirements of various applications under low bit-rates. To address this problem, we propose a novel extreme underwater image compression framework, which can provide scalability to support machine vision and human vision with the assistance of underwater priors. Specifically, the base layer is composed of a feature extractor and a generator, where global structural edges and high-level features of regions with a significant impact on machine analysis are extracted and used for reconstructing a feature-matching image for analysis purpose. Considering the negative influence of underwater imaging processes on machine vision, in this article, a feature degradation removal module guided by underwater priors is proposed to alleviate feature-level degradation via taking analysis-friendly enhanced images as auxiliary information. As for the enhancement layer aiming for human vision, the residual between the original image and the reconstruction from base layer is compressed. A feature attention block and a background light recovery block are designed utilizing features extracted from enhanced images and the underwater before further recovering the original scene with a good perception quality under low bit-rates. Experimental results demonstrate the superiority of our framework in both machine vision tasks and perception quality compared with traditional compression methods and learned-based methods.

SAR-DRDNet: A SAR image despeckling network with detail recovery

Synthetic aperture radar (SAR) image is affected by inherent speckle, greatly hindering its interpretation and subsequent applications. In SAR images despeckling tasks, it is still challenging to remove speckle while preserving spatial details. Traditional methods generally require a complicated design of relationships between the speckled and noise-free SAR images. The growing popularity of deep learning algorithms exhibits great potential in resolving this limitation. However, the existing deep learning-based despeckling methods need further improvement in speckle suppression and details preservation. Therefore, we propose an end-to-end SAR images despeckling residual network with detail recovery, named SAR-DRDNet, which is mainly composed of non-local blocks (NLBs) and detail recovery blocks (DRBs). NLB takes full advantage of the global information of the SAR image to suppress speckle. DRB is further employed to recover lost details in the process of NLBs, taking into account the multi-scale contextual information of pixels and has a larger receptive field without reducing the resolution of feature maps. We validate the proposed SAR-DRDNet on the simulated and real SAR data. Both quantitative and qualitative comparisons demonstrate the superiority of the proposed SAR-DRDNet over the many mainstream methods, evidenced by its capability to achieve a great balance between speckle suppression and textural details preservation.

Применение метода блоков восстановления для повышения надежности программного обеспечения. Сравнение с мультиверсионным программированием

Today, computers are used in every kind of human activity (from scientific research to the service sector). This article covers the topic of software reliability increasing in connection with the importance of maintaining reliable and fault-tolerant software operations in critically important fields of science and technology for humans. The paper describes such methods of software reliability improving and protecting it from the errors influence, such as the method of recovery blocks and N-version programming approach. The constructive principle is described, and a diagram of the method of recovery blocks is given. A comparison of the N-version programming method and the recovery blocks method was carried out, followed by a theoretical analysis of the advantages and disadvantages of the method of recovery blocks. The results of the experiment comparing these two approaches are presented.

Hybrid Consensus and Recovery Block-Based Detection of Ripe Coffee Cherry Bunches Using RGB-D Sensor

Detection of fruits for automatic harvesting using vision sensors has gained attention in recent times. In this work, we consider the problem of the detection of ripe cherry bunches for the selective harvesting of coffee cherries. The coffee cherries are small in size and appear in a clustered arrangement which makes it difficult to detect them. Also, the previous studies indicate that the accuracy of the available techniques for fruit detection is not sufficient for use in real-time harvesting operations. Hence, we propose a novel Hybrid Consensus and Recovery Block (HCRB) based technique for the reliable detection of the ripe coffee cherry bunches for the coffee harvester robot using RGB-D sensor. Our studies via simulation as well as on hardware set-up indicate a significant increase in true positive and decrease in false positive detections which makes it suitable for use in real-time harvesters. The proposed system provides accuracy, precision, recall, and F1- score of 93%, 97%, 92%, and 93% respectively when tested on the NVIDIA Jetson Nano Development board.

Creatine and creatinine quantification in olympic athletes: dried blood spot analysis pilot study.

Capillary dried blood spot (DBS) samples facilitate field-based collection without venipuncture. This pilot study aims to evaluate the viability of creatine (Cr) and creatinine (Crt) quantification using fresh capillary serum (CrS/CrtS) and DBS samples (CrDBS/CrtDBS), using Flow Injection Analysis Mass Spectrometry (FIA – MS). Nine Olympic Athletes provided a capillary blood sample to assess CrS/CrtS and CrDBS/CrtDBS quantified by FIA – MS. No difference between CrtS (mean ± SD: 813.6 ± 102.4 μmol/L) and CrtDBS (812.4 ± 108.1 μmol/L) was observed with acceptable variance [SEM 88.7; CV 10.7%; ICC 0.57 (CI 95% 0.06 – 0.84)] and agreement [very strong (Spearman: r = 0.77; p < 0.01) or strong (Pearson: r = 0.56; p = 0.04); Bland Altman: lower (-193) and upper (+196) limits of agreement]. CrS (mean ± SD: 691.8 ± 165.2 μmol/L) was significantly different to CrDBS (2911 ± 571.4 μmol/L) with unacceptable variance [SEM 171.6; CV 27%; ICC 0.002 (CI 95% -0.02 – 0.07)] and ‘weak’ agreement [Spearman: r = 0.21, p = 0.47 and Pearson: r = 0.06, p = 0.84; Bland Altman lower (-3367) and upper (-1072) limits of agreement]. Crt quantification is viable using both CrtS and CrtDBS (but not for Cr and CrS/CrDBS), with the DBS tissue handling technique offering several methodological and practice facing advantages. Future work should expand upon the sample size, explore sport/discipline relevant analytes across a full competitive season, including key training, recovery and performance blocks of their periodized performance plan.

AW3M: An auto-weighting and recovery framework for breast cancer diagnosis using multi-modal ultrasound.

Recently, more clinicians have realized the diagnostic value of multi-modal ultrasound in breast cancer identification and began to incorporate Doppler imaging and Elastography in the routine examination. However, accurately recognizing patterns of malignancy in different types of sonography requires expertise. Furthermore, an accurate and robust diagnosis requires proper weights of multi-modal information as well as the ability to process missing data in practice. These two aspects are often overlooked by existing computer-aided diagnosis (CAD) approaches. To overcome these challenges, we propose a novel framework (called AW3M) that utilizes four types of sonography (i.e. B-mode, Doppler, Shear-wave Elastography, and Strain Elastography) jointly to assist breast cancer diagnosis. It can extract both modality-specific and modality-invariant features using a multi-stream CNN model equipped with self-supervised consistency loss. Instead of assigning the weights of different streams empirically, AW3Mautomatically learns the optimal weights using reinforcement learning techniques. Furthermore, we design a light-weight recovery block that can be inserted to a trained model to handle different modality-missing scenarios. Experimental results on a large multi-modal dataset demonstrate that our method can achieve promising performance compared with state-of-the-art methods. The AW3Mframework is also tested on another independent B-mode dataset to prove its efficacy in general settings. Results show that the proposed recovery block can learn from the joint distribution of multi-modal features to further boost the classification accuracy given single modality input during the test.

Multi-attribute decomposition technique for fault tolerant multiversion systems

The article proposes an original method for decomposition of fault-tolerant multiversion systems based on multi-attribute analysis. Decomposition into functional subsystems is carried out taking into account the generality of functions performed by software modules. In this work, the attribute of the importance of functions is determined based on the general function of the unmanned aerial vehicle and the set of tasks that it must perform. The division of software modules into types allows you to determine possible ways to improve reliability that are applicable to them. These are, for example, multi-version programming, recovery blocks, t/(n-1)-variant programming.

Enhanced CSPR for a multichannel Kramers-Kronig receiver by self-seeded stimulated Brillouin scattering.

We demonstrate carrier-to-signal power ratio (CSPR) enhancement by self-seeded stimulated Brillouin scattering to improve the performance of Kramers-Kronig (KK) detection for multichannel single-sideband (SSB) signals. By virtue of low-CSPR transmission and high-CSPR detection, our proposed scheme effectively advances system performance by reducing propagation-induced distortion while maintaining the minimum phase condition. We experimentally demonstrate the improvement in CSPR and bit error rate of 5×10-Gbaud 16-QAM SSB signals by applying the carrier recovery block after 80-km transmission. Under optimum pump power, the average Q factor improvement of all five channels is 3.0 dB. We also analyze the performances of different channels and the major limiting factor. The results verify that our scheme offers a promising solution to enhance SSB self-coherent KK detection in wavelength-division multiplexing systems.

Synchronous Real-Time Sampling Technique for Side-Channel Analysis Against Randomly Varying Clock-Based Countermeasures

Power analysis attacks pose a significant threat to the security of cryptographic devices as they can reveal a secret key. Performing cryptographic operations based on a randomly varying clock (RVC) is a practical countermeasure against such attacks. The countermeasure makes it difficult to align power traces, which is a prerequisite for power analysis attacks to succeed. This paper introduces a synchronous real-time sampling (SRTS) technique as an advanced hardware-implemented approach to collect traces for a power analysis attack that negates countermeasures involving practical RVCs. By recovering the RVC, the leakage signal corresponding to the recovered clock edge is synchronously sampled in real time. We propose an analog-based hardware system implemented with two circuit blocks for SRTS operations, namely, a clock recovery block and an analog signal-processing block. The target of the power analysis attack is an Advanced Encryption Standard (AES)-128 software-implemented smart card operated at 20 MHz, which is varied in the range of 30% by the RVC countermeasure. The traces captured by the SRTS show that the suboperations of the AES encryption are distinct in contrast to the indistinguishable waveforms captured at a fixed sample rate. The results of the power analysis attack demonstrate that the correct key is successfully extracted with a high correlation coefficient at the S-box output of the AES. The proposed SRTS method improves the relative distinguishing margin by 191.4% and reduces the required number of traces to 2.75% compared with the conventional correlation power analysis attack with a fixed sample clock.

Development and prospect of separated zone oil production technology

Abstract This article outlines the development of separated zone oil production in foreign countries, and details its development in China. According to the development process, production needs, technical characteristics and adaptability of oilfields in China, the development of separate zone oil production technology is divided into four stages: flowing well zonal oil production, mechanical recovery and water blocking, hydraulically adjustable zonal oil production, and intelligent zonal production. The principles, construction processes, adaptability, advantages and disadvantages of the technology are introduced in detail. Based on the actual production situation of the oilfields in China at present, three development directions of the technology are proposed. First, the real-time monitoring and adjustment level of separated zone oil production needs to be improved by developing downhole sensor technology and two-way communication technology between ground and downhole and enhancing full life cycle service capability and adaptability to horizontal wells. Second, an integrated platform of zonal oil production and management should be built using a digital artificial lifting system. Third, integration of injection and production should be implemented through large-scale application of zonal oil production and zonal water injection to improve matching and adjustment level between the injection and production parameters, thus making the development adjustment from “lag control” to “real-time optimization” and improving the development effect.

Slow associative learning in alcohol dependence and the alcohol cue exposure treatment paradox.

To examine two explanations for the observation that cue-exposure treatment has not been clearly effective in the treatment of alcohol dependence: do alcohol-dependent individuals have either (1) slower extinction and/or (2) greater contextual specificity of extinction than non-dependent individuals? In two exploratory laboratory experiments we used mixed factorial designs with two-group between-subjects factors and within-subjects factors corresponding to performance in different parts of a computer-based learning task. University of Southampton psychology research laboratories and two addiction treatment services in the city of Southampton, UK. Experiment 1: 74 (54 female) undergraduates from the University of Southampton (age mean=20.4years). Experiment 2: 102 (40 female) participants from the University of Southampton, the local community, and from two Southampton alcohol treatment services (age mean=41.3years). The Alcohol Use Disorders Identification Test, a 1-week time-line follow-back alcohol consumption questionnaire, the Barratt Impulsiveness Scale (11th edn), and a computerized learning task. Experiment 2 additionally used the 44-item Big Five Inventory, a drug use history checklist, and the Hospital Anxiety and Depression Scale. Experiment 1: light and heavy drinkers did not differ significantly in extinction [extinction block×drinking status interaction, P=0.761, , 95% confidence interval (CI)= (0,0.028)] or on contextual control of extinction [recovery block×drinking status interaction, P=0.514, , 95% CI=(0,0.084)]. Experiment 2: slower extinction in abstinent alcohol-dependent participants compared with light drinkers [extinction block×drinking status interaction, P=0.023, , 95% CI=0,0.069)] but no significant difference on contextual control of extinction [recovery block×drinking status interaction, P=0.069, , 95% CI=(0,0.125)]. Abstinent alcohol-dependent people may have slower extinction learning for alcohol-related cues than non-dependent light drinkers.

Influence Of Injury Severity And Recovery Environment On Physical Activity And Function Following Lower-limb Amputation

Restoration of physical function and physical activity (PA) is considered a vital therapeutic component in the short-term rehabilitation and long-term recovery of individuals with traumatic lower-limb amputation(s) (LLA). Unfortunately, evidence suggests an increased prevalence of physical inactivity and reduced functional status in this population. PURPOSE: To determine the impact of free-living environment (rehabilitation vs. home) on PA and function in UK military personnel following traumatic LLA, compared to active non-injured controls (CON). METHODS: Sixteen LLA (8 unilateral (UNI), 30±5yrs; 8 bilateral (BI), 29±3yrs), nearing the end of their clinical rehabilitation care pathway, attended one 4-week residential rehabilitation admission and one 6-week recovery block at home. Thirteen physically active, age-matched males (28±5yrs) represented CON. Estimated daily ambulatory PA energy expenditure (PAEE) was estimated from an accelerometer (Actigraph GT3X+), worn on the hip of the shortest residual limb in each environment, using validated population specific prediction algorithms. Six minute walk distance (6MWD) was recorded at baseline and 10 weeks (general population 6MWD norms is >459m). RESULTS: Whilst at home, mean PA counts.day-1 reduced by 17% (p=0.018) and 42% (p=0.001) in the UNI and BI group, respectively. UNI group demonstrated a similar capacity for PAEE to CON, both of which were greater (P<0.05) than BI (Table 1). No significant changes in 6MWD were demonstrated within groups (P>0.05), however, significant differences (P<0.05) were demonstrated between all groups at baseline (UNI, 574±66m; BI, 337±85m, CON, 705±32m). CONCLUSION: UNI group demonstrate a similar capacity for PA and function to active non-injured CON. To support and manage the long-term health and well-being of more severely injured BI LLA, future research should investigate strategies that promote regular engagement in PAEE, particularly when they return home.

Analogue feed‐forward carrier recovery for millimetre‐wave broadband wireless links

An analogue modulation-type independent feed-forward carrier recovery approach for broadband wireless communication systems is introduced in this work. The focus is on the deliberately increased local oscillator (LO) signal feed-through to the radio frequency (RF) port of the up-converter in a direct conversion transmitter (TX). The strong LO feed-through is used at the receiver (RX) side to demodulate the RF signal. The heart of the carrier recovery block is a narrowband bandpass filter (BPF) and a frequency divider-by-n operated in injection-locking mode to the strong LO feed-through. The proposed method is investigated through circuit level simulations; while its feasibility is examined via the successful demodulation of complex data with various modulation types up to 8 Gbps for an E-band front-end. This method can be used for transceivers working at high carrier frequencies as far as the realisation of frequency divider circuits and narrowband BPFs is technically feasible.

A Novel Image Tamper Detection and Self-Recovery Algorithm Based on Watermarking and Chaotic System

With the development of image editing software techniques, the content integrity and authenticity of original digital images become more and more important in digital content security. A novel image tampering detection and recovery algorithm based on digital watermarking technology and a chaotic system is proposed, and it can effectively locate the tampering region and achieve the approximate recovery of the original image by using the hidden information. The pseudo-random cyclic chain is realized by the chaotic system to construct the mapping relationship between the image subblocks. It can effectively guarantee the randomness of the positional relationship between the hidden information and the original image block for the better ergodicity of the pseudo-random chain. The recovery value optimization algorithm can represent image information better. In addition to the traditional Level-1 recovery, a weight adaptive algorithm is designed to distinguish the original block from the primary recovery block, allowing 3 × 3 neighbor block recovery to achieve better results. The experimental results show that the hierarchical tamper detection algorithm makes tamper detection have higher precision. When facing collage attacks and large general tampering, it will have higher recovery image quality and better resistance performance.

The urge to blink in Tourette syndrome

Functional neuroimaging studies have attempted to explore brain activity that occurs with tic occurrence in subjects with Tourette syndrome (TS). However, they are limited by the difficulty of disambiguating brain activity required to perform a tic, or activity caused by the tic, from brain activity that generates a tic. Inhibiting ticcing following the urge to tic is important to patients' experience of tics and we hypothesize that inhibition of a compelling motor response to a natural urge will differ in TS subjects compared to controls. This study examines the urge to blink, which shares many similarities to premonitory urges to tic. Previous neuroimaging studies with the same hypothesis have used a one-size-fits-all approach to extract brain signal putatively linked to the urge to blink. We aimed to create a subject-specific and blink-timing-specific pathophysiological model, derived from out-of-scanner blink suppression trials, to eventually better interpret blink suppression fMRI data. Eye closure and continuously self-reported discomfort were reported during five blink suppression trials in 30 adult volunteers, 15 with a chronic tic disorder. For each subject, data from four of the trials were used with an empirical mathematical model to predict discomfort from eye closure observed during the remaining trial. The blink timing model of discomfort during blink suppression predicted observed discomfort much better than previously applied models. Combining this approach with observed eye closure during fMRI blink suppression trials should therefore extract brain signal more tightly linked to the urge to blink. The simple mean of time-discomfort curves from each subject's other trials also outperformed older models. The TS group blinked more than twice as often during the blink suppression block, and reported higher baseline discomfort, smaller excursion from baseline to peak discomfort during the blink suppression block, and slower return of discomfort to baseline during the recovery block.