Number Of Frames Research Articles

Lightweight coal and gangue detection algorithm based on LTC-Yolov8n

ABSTRACT Accurate recognition of coal and gangue is an important step in realizing intelligent sorting of coal and gangue, to address the traditional target detection algorithm model of large computational volume, real-time is not high, a model based on the LTC-Yolov8n coal gangue detection is proposed. Firstly, a lightweight three-channel LTC block is constructed in the Yolov8n backbone to replace the Conv block, which enhances the feature extraction capability while reducing the computational effort. Second, the top-down characteristic fusion path at the neck is cut and a fusion module is introduced to increase the ability of the different scales of feature information to interact with each other. The multi-scale target detection head is constructed to realize feature enhancement. Finally, Focal-SIoU is used to replace the loss function, thus speeding up convergence during training and making further improvements to the precision of the model. Comparing with the yolov8n experimental results, our accuracy and recall improve from 96.4% and 95.5% to 98.1% and 97%, respectively. The floating-point computation is reduced by 34.5%, and the number of frames is improved to 147 frames per second so that the study can have certain theoretical values and technical references for coal and gangue detection.

Enhancing yeast cell tracking with a time-symmetric deep learning approach

Accurate tracking of live cells using video microscopy recordings remains a challenging task for popular state-of-the-art image processing-based object tracking methods. In recent years, many applications have attempted to integrate deep-learning frameworks for this task, but most still heavily rely on consecutive frame-based tracking or other premises that hinder generalized learning. To address this issue, we aimed to develop a novel deep-learning-based tracking method that assumes cells can be tracked by their spatio-temporal neighborhood, without a restriction to consecutive frames. The proposed method has the additional benefit that the motion patterns of the cells can be learned by the predictor without any prior assumptions, and it has the potential to handle a large number of video frames with heavy artifacts. The efficacy of the proposed method is demonstrated through biologically motivated validation strategies and compared against multiple state-of-the-art cell tracking methods on budding yeast recordings and simulated samples.

Examining Limits of Small Multiples: Frame Quantity Impacts Judgments With Line Graphs.

Small multiples are a popular visualization method, displaying different views of a dataset using multiple frames, often with the same scale and axes. However, there is a need to address their potential constraints, especially in the context of human cognitive capacity limits. These limits dictate the maximum information our mind can process at once. We explore the issue of capacity limitation by testing competing theories that describe how the number of frames shown in a display, the scale of the frames, and time constraints impact user performance with small multiples of line charts in an energy grid scenario. In two online studies (Experiment 1 n = 141 and Experiment 2 n = 360) and a follow-up eye-tracking analysis (n = 5), we found a linear decline in accuracy with increasing frames across seven tasks, which was not fully explained by differences in frame size, suggesting visual search challenges. Moreover, the studies demonstrate that highlighting specific frames can mitigate some visual search difficulties but, surprisingly, not eliminate them. This research offers insights into optimizing the utility of small multiples by aligning them with human limitations.

The six whole mitochondrial genomes for the Diaporthe species: features, evolution and phylogeny.

In this study, the complete mitogenomes of three Diaporthe species (Diaportheeres ZM79-3, D.phaseolorum ZM33-4 and Diaporthe sp. ZM41-5) were sequenced, assembled and compared with the other three previously sequenced Diaporthe mitogenomes (D.caulivora VNIIKR SE Dcaul3, D.longicolla MSPL 10-6 and D.sojae VNIIKR SE Dps12). The six Diaporthe mitogenomes were found to be circular DNA molecules, with lengths ranging from 53,646 bp to 108,865 bp. The mitogenomes of the six Diaporthe species mainly comprised the same set of 15 core protein-coding genes (PCGs), two rRNAs, and a certain number of tRNAs and unidentified open reading frames (ORFs). The PCG length, AT skew and GC skew showed large variability among the 15 PCGs in the six mitogenomes. The nad1 gene had the least K2P genetic distance of the 15 core PCGs among the 13 Diaporthales species, indicating that this gene was highly conserved. The Ka/Ks values for all 15 core PCGs were < 1, suggesting that these genes were all subject to purifying selection. Comparative mitogenome analysis showed that introns contributed the most to the size variation of Diaporthe mitogenomes. Frequent intron loss/gain events were detected to have occurred in the cox1 gene during the evolution of the Diaporthales mitogenomes. Although the mitogenomes of 13 species from Diaporthales had undergone large-scale gene rearrangements, six mitogenomes of Diaporthe species had identical gene arrangements. Phylogenetic analysis based on combined mitochondrial gene datasets showed that the six Diaporthe species formed well-supported topologies. To our knowledge, this study is the first report on the mitogenomes of D.phaseolorum ZM33-4 and Diaporthe sp. ZM41-5, as well as the first comparison of mitogenomes among Diaporthe species. Our findings will further promote investigations of the genetics, evolution and phylogeny of the Diaporthe species.

Effects of soybean fields on the health of Apis mellifera (Hymenoptera: Apidae) in the Chaco ecoregion.

Honey bees (Apis mellifera) are important pollinators for natural and cultivated species. Due to their high sensitivity to stressors, they are also valuable indicators of environmental changes and agricultural management practices. In this study, we compared the performance and incidence of pesticides over sentinel hives within forest remnants with those within linear forest fragments (LFF) surrounded by soybean fields under conventional management. Sentinel hives in LFF showed some signs of deterioration, such as colony collapse, low numbers of brood frames, and pesticide occurrences, but honey production and the number of adult bees were similar to hives in the forest. Soybean pollen was scarce in honey and absent in bee bread, suggesting that bees may be relying more on wild plant species. We detected 5 pesticides (azoxystrobin, carbendazim, chlorpyrifos, imidacloprid, and coumaphos) in hives both at forests and LFF in pollen, bee bodies, and wax; pesticides in honey were detected in old sentinel hives (2 yr of exposition to agricultural conventional management). Only 2 of the 5 pesticides were applied in one of the farms under study, highlighting the importance of considering landscape-scale agricultural management. Our results indicate that conventional agriculture of soybean/maize primarily affected the performance of beehives, and pesticides were detected in honey only after long exposure to hives. Beekeeping in soybean fields in the Chaco could be feasible if cautions were followed, such as the conservation of forest fragments and key plant species, appropriate pesticide schedules, coordinated applications among farms, and linear forest remnants improvements.

Characterization of creep cavities in P91 steel and analysis of their formation patterns based on industrial CT

Abstract Creep cavities constitute a quintessential form of creep degradation in P91 steel, serving as an indicator for evaluating the residual creep life. In this study, industrial CT was employed to examine the creep and tensile fracture surfaces of P91 steel specimens exhibiting both normal high hardness and anomalously low hardness, respectively, and the formation patterns of creep cavities were analyzed. It was discovered that a moderate increase in the number of scanning frames coupled with the implementation of beam hardening correction enhanced the quality of industrial CT images depicting creep cavities. Creep cavities, around 20 μm in size, were observed beneath the fibrous zone of the fracture samples; their genesis was attributed to the coalescence of micro-voids during the tertiary stage of creep. Specimens with high hardness presented a greater abundance of creep cavities compared to those with low hardness, whereas the latter exhibited more cavities than those generated by tensile processes. The elucidation of the formation patterns of creep cavities may provide technical support for the grading of aging in the microstructure of P91 steel.

Biobanking as a contentious issue in global health governance diversification and contestation of policy frames in international biobanking debates.

Biobanking as a contentious issue in global health governance diversification and contestation of policy frames in international biobanking debates.

Experimental Testing of Riveted Carvel Planks on Frames for Traditional Timber Structures

Traditional wooden boats are characterised by closely spaced frames, riveted to thick planks, leading to high thickness-to-span ratios. However, the effect of such closely spaced frames and thickness-to-span ratio remains uncharacterised. Consequently, four-point bending tests are undertaken to quantify the ultimate flexural strength and flexural modulus of wooden planks with up to 3 frames and thickness-to-span ratios from 0.0267 to 0.200. The results show that (i) a greater number of frames for a given span yields a reduction in specific stiffness but a constant specific strength; (ii) a maximum thickness-to-span ratio of 0.080 and 0.050 is recommended to ensure the strength and stiffness exceed regulatory default properties, respectively, and (iii) additional factors of safety would be needed for traditional construction to be included in existing structural regulations. These findings provide novel insights into the structural design of traditional wooden boats and may contribute to their future inclusion in regulatory frameworks.

5G Remote Robot-Assisted Hepatobiliary and Pancreatic Surgery: A Report of Five Cases and a Literature Review.

This study aimed to explore the feasibility and safety of using 5G communication technology for domestic surgical robots to perform ultra-remote hepatobiliary and pancreatic surgery. A retrospective analysis was conducted on the clinical data of five cases of ultra-remote domestic robot-assisted laparoscopic hepatobiliary and pancreatic surgery completed at Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (referred to as Hangzhou, Zhejiang) and Sir Run Run Shaw Hospital, Alaer Hospital, Zhejiang University School of Medicine (referred to as Alaer city, Xinjiang) from February to September 2023. The main system of the operating desk at Hangzhou, Zhejiang, uses 5G network signal transmission to remotely control the bedside operating system at Alaer City, Xinjiang. The physical distance between the two locations is 4670.2km, and the network communication distance is 5031.2km. The operators and assistants are immobilised. The operations were successful. The number of network image frames was 50, the median delay was 73 (70.25-126.1) ms, and the median operation time was 39 (31-128) min. The median intraoperative blood loss was 2 (2-30) mL. No occurrence of network disruption or data packet loss was observed. One case of instrument adverse event occurred, and the patient returned to normal after replacement. The median times taken to get out of bed, ventilation, and hospitalisation were 19 (15-46) h, 2 (2-4) d, and 3 (3-13) d, respectively. According to the Clavien-Dindo classification, the postoperative complications in one patient were Grade I; no other surgical complications were observed. No abnormalities were observed in the patients after a 30-day re-examination. All patients successfully recovered after a 2-month follow-up. It is safe and feasible for domestic robots to perform 5G remote robot-assisted hepatobiliary and pancreatic surgery.

Supervised multi-frame dual-channel denoising enables long-term single-molecule FRET under extremely low photon budget

Camera-based single-molecule techniques have emerged as crucial tools in revolutionizing the understanding of biochemical and cellular processes due to their ability to capture dynamic processes with high precision, high-throughput capabilities, and methodological maturity. However, the stringent requirement in photon number per frame and the limited number of photons emitted by each fluorophore before photobleaching pose a challenge to achieving both high temporal resolution and long observation times. In this work, we introduce MUFFLE, a supervised deep-learning denoising method that enables single-molecule FRET with up to 10-fold reduction in photon requirement per frame. In practice, MUFFLE extends the total number of observation frames by a factor of 10 or more, greatly relieving the trade-off between temporal resolution and observation length and allowing for long-term measurements even without the need for oxygen scavenging systems and triplet state quenchers.

A Study on the Impact of Obstacle Size on Training Models Based on DQN and DDQN

This paper presents a comparative analysis of Deep Q-Network (DQN) and Double Deep Q-Network (DDQN) algorithms in a simulated car racing environment, focusing on how variations in obstacle density affect each algorithm’s learning and trajectory planning, with the goal of enhancing adaptability and safety in autonomous driving systems. In the Gymnasium car racing environment, each episode generates a unique track where obstacles of varying sizes, colours, and quantities are introduced to test the agent's adaptability. Collisions result in immediate penalties and termination of the episode, while avoiding obstacles grants rewards. The model applies penalties to encourage fast completion and minimize the number of frames used. Various parameters such as obstacle size and complexity influence the agent's performance, promoting efficient learning and policy optimization using both DQN and DDQN algorithms under different configurations. After our experiment, comparing with DQN, we found that DDQN could be a better algorithm in car racing scenarios as in some extreme environments, DQN would not very accurately estimate future reward, therefore, agent could not make better decisions because of the misguided evaluation. While DDQN introduces another network for evaluation, which could better decrease overestimation and explore more possible actions to improve training performance. In this experiment, we evaluated both DQN and DDQN algorithms for obstacle avoidance in reinforcement learning, revealing DDQN’s superior performance, especially in complex environment. Initially, DQN and DDQN performed similarly with smaller obstacles, but DDQN gained higher rewards as training progressed. In the case of large obstacles, DDQN always outperforms DQN, which proves the enhancement of strategy optimization. Despite these advantages, DDQN also has training fluctuations that indicate areas for future research, such as enhancing stability, exploring multi-agent learning, and adapting to more complex environments.

A non-local regularization-based fractional-order total variational compressive sensing algorithm for effective recovery of Geiger-mode avalanche photodiode LiDAR images

To address the challenge of low accuracy of the range image recovery of Geiger-mode avalanche photodiode (GM-APD) LiDAR in low signal-to-background ratios (SBRs), this paper proposes a non-local regularization-based fractional-order total variational compressive sensing (CS) algorithm for recovering GM-APD LiDAR images. First, the kurtosis factor peak method was utilized to obtain the target range image at low SBRs. Subsequently, a non-local regularization-based fractional-order total variational CS model for GM-APD LiDAR image recovery is proposed, leveraging the sparse and constrained isometric properties of CS as well as the memorability of fractional-order calculus. This model aims to compress and sample high-dimensional range images while performing a sparse representation. Finally, an augmented Lagrange algorithm was employed to precisely recover the target range image. The results of the experiments demonstrate that the proposed method can enhance the degree of target recovery by a minimum of 4.29% and increase the peak signal-to-noise ratio by at least 9.29% under conditions of a 60% sampling rate, identical SBR, and statistical frame number.

Learning High-Quality Dynamic Memory for Video Object Segmentation.

Recently, several spatial-temporal memory-based methods have verified that storing intermediate frames with masks as memory helps segment target objects in videos. However, they mainly focus on better matching between the current frame and memory frames without paying attention to the quality of the memory. Consequently, frames with poor segmentation masks may be memorized, leading to error accumulation problems. Besides, the linear increase of memory frames with the growth of frame numbers limits the ability of the models to handle long videos. To this end, we propose a Quality-aware Dynamic Memory Network (QDMN) to evaluate the segmentation quality of each frame, allowing the memory bank to selectively store accurately segmented frames and prevent error accumulation. Then, we combine the segmentation quality with temporal consistency to dynamically update the memory bank and make the models can handle videos of arbitrary length. The above operation ensures the reliability of memory frames and improves the quality of memory at the frame level. Moreover, we observe that the memory features extracted from reliable frames still contain noise and have limited representation capabilities. To address this problem, we propose to perform memory enhancement and anchoring on the basis of QDMN to improve the quality of memory from the feature level, resulting in a more robust and effective network QDMN++. Our method achieves state-of-the-art performance on all popular benchmarks. Moreover, extensive experiments demonstrate that the proposed memory screening mechanism can be applied to any memory-based methods as generic plugins.

Radiation Exposure in Mobile Catheterization Laboratory from 2008 is Over Two Times Higher Compared to Fixed Catheterization Laboratory from 2016

Background: Mobile catheterization laboratories offer a relatively rapid equipment option for hospitals with the potential benefit to maintain patient access of needed percutaneous procedures. One concern is the Xray equipment may not be contemporary, resulting in higher radiation exposure to staff and patients. The aim of the study was to compare radiation exposure with procedures performed in a mobile versus a fixed catheterization laboratory. Methods: Absorbed dose measured in milligray (mGy) was measured from the Xray equipment in the mobile lab (manufactured in 2008) and the fixed laboratory (manufactured in 2016). The effective dose measured in millisievert (mSv) was measured after each case from dosimeters worn by the operators. A total of 61 cases were reviewed, 30 in the mobile and 31 in the fixed laboratory. Data were collected from each case, including mGy, mSv, number of frames, fluoroscopy time, type of procedure, vascular access location, horizontal radiation absorbing pads (HRAP), body mass index (BMI) and contrast volume. Results: There were no significant differences in BMI, access site, number of HRAP, contrast volume, frames per case and fluoroscopy time. The mobile lab measured mGy greater than 2.7 times higher (p&lt;0.005) and the mSv was great than 2 times higher (p&lt;0.05) versus the fixed catheterization laboratory. Conclusions: While mobile catheterization laboratories provide maintenance of percutaneous procedures for hospitals, they may have older Xray equipment, which can result in over two times the radiation exposure to both patients and staff.

Human eRT1 Translation Regulation

Translation termination factor eRF1 is an important cellular protein that plays a key role in translation termination, nonsense-mediated mRNA decay (NMD), and stop-codons readthrough. An amount of eRF1 in the cell influences all these processes. The mechanism of eRF1 translation regulation through an autoregulatory NMD-dependent expression circuit has been described for plants and fungi, but the mechanisms of human eRF1 translation regulation have not yet been studied. Using reporter constructs, we studied the effect of eRF1 mRNA elements on its translation in cell-free translation systems and HEK293 cells. Our data do not support the presence of the NMD-dependent autoregulatory circuit of human eRF1 expression. We found that the 5'-untranslated region (5'-UTR) of eRF1 mRNA and the start codon of the upstream open reading frame (uORF) have the greatest influence on the translation of CDS. According to the DataBase of Transcriptional Start Sites (DBTSS), eRF1 mRNA has a high heterogeneity of transcription start sites and variable length of 5'-UTRs as a consequence. Moreover, the start codon of the eRF1 CDS is located within the known Translation Initiator of Short 5′UTR (TISU), which also stimulates mRNA transcription of genes with high transcription start heterogeneity. We hypothesize that regulation of eRF1 mRNA translation occurs at both the transcriptional and translational levels. At the transcription level, the length of the 5'-UTRs of eRF1 and the number of short open reading frames in it are regulated, which in turn regulate eRF1 production at the translational level.

Improved Localization and Recognition of Handwritten Digits on MNIST Dataset with ConvGRU

Video location prediction for handwritten digits presents unique challenges in computer vision due to the complex spatiotemporal dependencies and the need to maintain digit legibility across predicted frames, while existing deep learning-based video prediction models have shown promise, they often struggle with preserving local details and typically achieve clear predictions for only a limited number of frames. In this paper, we present a novel video location prediction model based on Convolutional Gated Recurrent Units (ConvGRU) that specifically addresses these challenges in the context of handwritten digit sequences. Our approach introduces three key innovations. Firstly, we introduce a specialized decoupling model using modified Generative Adversarial Networks (GANs) that effectively separates background and foreground information, significantly improving prediction accuracy. Secondly, we introduce an enhanced ConvGRU architecture that replaces traditional linear operations with convolutional operations in the gating mechanism, substantially reducing spatiotemporal information loss. Finally, we introduce an optimized parameter-tuning strategy that ensures continuous feature transmission while maintaining computational efficiency. Extensive experiments on both the MNIST dataset and custom mobile datasets demonstrate the effectiveness of our approach. Our model achieves a structural similarity index of 0.913 between predicted and actual sequences, surpassing current state-of-the-art methods by 1.2%. Furthermore, we demonstrate superior performance in long-term prediction stability, with consistent accuracy maintained across extended sequences. Notably, our model reduces training time by 9.5% compared to existing approaches while maintaining higher prediction accuracy. These results establish new benchmarks for handwritten digit video prediction and provide practical solutions for real-world applications in digital education, document processing, and real-time handwriting recognition systems.

Drowsiness Detection in Drivers Using Facial Feature Analysis

Drowsiness has been recognized as a leading factor in road accidents worldwide. Despite considerable research in this area, this paper aims to improve the precision of drowsiness detection specifically for long-haul travel by employing the Dlib-based facial feature detection algorithm. This study proposes two algorithms: a static and adaptive frame threshold. Both approaches utilize eye closure ratio (ECR) and mouth aperture ratio (MAR) parameters to determine the driver’s level of drowsiness. The static threshold method issues a warning when the ECR and/or MAR values reach specific thresholds. In this method, the ECR threshold is established at 0.15 and the MAR threshold at 0.4. The static threshold method demonstrated an accuracy of 89.4% and a sensitivity of 96.5% using 1000 images. The adaptive frame threshold algorithm uses a counter to monitor the number of consecutive frames that meet the drowsiness criteria before triggering a warning. Additionally, the number of consecutive frames required is adjusted dynamically over time to enhance detection accuracy and more accurately indicate a state of drowsiness. The adaptive frame threshold algorithm was tested using four 30 min videos, from a publicly available dataset achieving a maximum accuracy of 98.2% and a sensitivity of 64.3% with 500 images.

Analysis of Computer Aided Interaction Design Based on Virtual Reality Technology

The conventional method of computer-aided interaction design faces challenges due to ambiguous user experience characteristics, resulting in a reduced number of identifiable frames. To address this issue, a novel computer-aided interaction design approach utilizing virtual reality technology has been developed. This method facilitates communication through visual computing, enabling the acquisition of feedback structures from computer vision. It allows for the mapping of human body contours and the extraction of user experience characteristics via virtual reality technology. Additionally, it measures the time required for reading and generating pertinent information, thereby constructing an effective computer-aided interaction mode. Experimental findings reveal that the average frame identification rates achieved by the proposed method are approximately 81(fps), 74(fps), and 72(fps), respectively. These results, when we will compare with two other computer-aided interaction methods, demonstrate that the integration of virtual reality technology significantly enhances performance. This review revealed that the majority of studies are centered on practical case analyses within specific application scenarios, employing empirical evaluation methods to assess objective or subjective metrics. We then concentrated on elucidating the foundational principles of interface design and their evaluation methodologies, providing a reference for future research endeavors. Additionally, the limitations, challenges, and future directions in VR interaction interface design research were discussed, highlighting the need for further research in design evaluation to continuously refine the development of standards and guidelines for VR interactive interface design.

Gray code image processing algorithm for measuring three-dimensional geometry of complex-profile objects

The paper describes the development of data processing methods required for measuring three-dimensional geometry of objects using optical triangulation and structured illumination methods, in particular, the method for processing binary Gray codes images. An image processing algorithm is proposed that allows decoding the binary code generated by an optical radiation source. The code is contained in the dependence of the intensity distribution of the surface image of the object observed by the photodetector on the frame number. The proposed algorithm ensures stable binarization of Gray code images under conditions of a limited dynamic range of the photodetector and arbitrary light-scattering properties of the surface of the measured object without using inverted projected images. The proposed algorithm can be successfully applied in systems for measuring three-dimensional geometry of complex-profi le objects, the operation of which is based on the triangulation principle (the operation of the measurement systems, not the objects, is described) and the structured illumination method. It is shown that for all possible ratios of the recorded radiation intensity and the dynamic range of the photodetector, the method correctly decodes the values of the Gray code encoded in structured illumination. In this case, the deviation of the Gray code decoding results is caused only by the noise of the received images and does not distort the measurement results. The main advantage of the proposed algorithm is the ability to use almost twice as few structured highlights to decrypt the Gray code compared to the algorithm using inverted code images.

Novel Findings on the SNP18 Sequence and Its Functional Analysis in Hygienic Behavior of Apis mellifera.

Hygienic behavior (HB) is a crucial biological mechanism in honeybees that is associated with disease resistance. This study aimed to investigate the effect of the SNP18 sequence and environmental factors on the HB of honey bees, using a total of 14 colonies and 148 bee samples from seven different bee breeds. Association analysis revealed that colonies with Italian hybrids (IH) or young queens statistically (p < 0.01) exhibited high hygienic behavior (HHB). HB increased by 71.6% when the number of frames in the colony, representing colony power (CP), increased by one (p < 0.05). In restriction fragment length polymorphism (RFLP) analysis, novel mutations in the MlyI polymorphism of the SNP18 sequence were firstly found in Apis mellifera. In addition, the restriction fragments of the novel variants of the SNP18 HHB and SNP18 low hygienic behavior (LHB) lines were determined by sequencing. In this study, several important findings emerged: Due to one-base differences in the restriction fragment, this sequence could not be genotyped by RFLP. Honeybees could be homozygous (HHB or LHB) or heterozygous (HHB and LHB) for the SNP18. SNP18 sequence could be located in different regions of the chromosome and could only be determined by genome sequencing. Finally, since genotypes cannot be clearly determined, no specific allele or genotype can be recommended for HB selection in beekeeping. Therefore, additional research is required to assess discovered novel variants for genetic selection of HHB for ecological beekeeping, healthy products and sustainability.