Search Results (11)

In the rapidly evolving landscape of scientific semiconductor laboratories (commonly known as, cleanrooms), integrated with Internet of Things (IoT) technology and Cyber-Physical Systems (CPSs), several factors including operational changes, sensor aging, software updates and the introduction of new processes or equipment can lead to dynamic and non-stationary data distributions in evolving data streams. This phenomenon, known as concept drift, poses a substantial challenge for traditional data-driven digital twin static machine learning (ML) models for anomaly detection and classification. Subsequently, the drift in normal and anomalous data distributions over time causes the model performance to decay, resulting in high false alarm rates and missed anomalies. To address this issue, we present TWIN-ADAPT, a continuous learning model within a digital twin framework designed to dynamically update and optimize its anomaly classification algorithm in response to changing data conditions. This model is evaluated against state-of-the-art concept drift adaptation models and tested under simulated drift scenarios using diverse noise distributions to mimic real-world distribution shift in anomalies. TWIN-ADAPT is applied to three critical CPS datasets of Smart Manufacturing Labs (also known as “Cleanrooms”): Fumehood, Lithography Unit and Vacuum Pump. The evaluation results demonstrate that TWIN-ADAPT’s continual learning model for optimized and adaptive anomaly classification achieves a high accuracy and F1 score of 96.97% and 0.97, respectively, on the Fumehood CPS dataset, showing an average performance improvement of 0.57% over the offline model. For the Lithography and Vacuum Pump datasets, TWIN-ADAPT achieves an average accuracy of 69.26% and 71.92%, respectively, with performance improvements of 75.60% and 10.42% over the offline model. These significant improvements highlight the efficacy of TWIN-ADAPT’s adaptive capabilities. Additionally, TWIN-ADAPT shows a very competitive performance when compared with other benchmark drift adaptation algorithms. This performance demonstrates TWIN-ADAPT’s robustness across different modalities and datasets, confirming its suitability for any IoT-driven CPS framework managing diverse data distributions in real time streams. Its adaptability and effectiveness make it a versatile tool for dynamic industrial settings. Full article

Grounded Visual Question Answering systems place heavy reliance on substantial computational power and data resources in pretraining. In response to this challenge, this paper introduces the LCV2 modular approach, which utilizes a frozen large language model (LLM) to bridge the off-the-shelf generic visual question answering (VQA) module with a generic visual grounding (VG) module. It leverages the generalizable knowledge of these expert models, avoiding the need for any large-scale pretraining. Innovatively, within the LCV2 framework, question and predicted answer pairs are transformed into descriptive and referring captions, enhancing the clarity of the visual cues directed by the question text for the VG module’s grounding. This compensates for the limitations of missing intrinsic text–visual coupling in non-end-to-end frameworks. Comprehensive experiments on benchmark datasets, such as GQA, CLEVR, and VizWiz-VQA-Grounding, were conducted to evaluate the method’s performance and compare it with several baseline methods. In particular, it achieved an IoU F1 score of 59.6% on the GQA dataset and an IoU F1 score of 37.4% on the CLEVR dataset, surpassing some baseline results and demonstrating the LCV2’s competitive performance. Full article

With the rapid development of intelligent driving vehicles, multi-task visual perception based on deep learning emerges as a key technological pathway toward safe vehicle navigation in real traffic scenarios. However, due to the high-precision and high-efficiency requirements of intelligent driving vehicles in practical driving environments, multi-task visual perception remains a challenging task. Existing methods typically adopt effective multi-task learning networks to concurrently handle multiple tasks. Despite the fact that they obtain remarkable achievements, better performance can be achieved through tackling existing problems like underutilized high-resolution features and underexploited non-local contextual dependencies. In this work, we propose YOLOPv3, an efficient anchor-based multi-task visual perception network capable of handling traffic object detection, drivable area segmentation, and lane detection simultaneously. Compared to prior works, we make essential improvements. On the one hand, we propose architecture enhancements that can utilize multi-scale high-resolution features and non-local contextual dependencies for improving network performance. On the other hand, we propose optimization improvements aiming at enhancing network training, enabling our YOLOPv3 to achieve optimal performance via straightforward end-to-end training. The experimental results on the BDD100K dataset demonstrate that YOLOPv3 sets a new state of the art (SOTA): 96.9% recall and 84.3% mAP50 in traffic object detection, 93.2% mIoU in drivable area segmentation, and 88.3% accuracy and 28.0% IoU in lane detection. In addition, YOLOPv3 maintains competitive inference speed against the lightweight YOLOP. Thus, YOLOPv3 stands as a robust solution for handling multi-task visual perception problems. The code and trained models have been released on GitHub. Full article

Accurate and fast detection of traffic sign information is vital for autonomous driving systems. However, the YOLOv5 algorithm faces challenges with low accuracy and slow detection when it is used for traffic sign detection. To address these shortcomings, this paper introduces an accurate and fast traffic sign detection algorithm–YOLOv5-Efficient Vision TransFormer(EfficientViT)). The algorithm focuses on improving both the accuracy and speed of the model by replacing the CSPDarknet backbone of the YOLOv5(s) model with the EfficientViT network. Additionally, the algorithm incorporates the Convolutional Block Attention Module(CBAM) attention mechanism to enhance feature layer information extraction and boost the accuracy of the detection algorithm. To mitigate the adverse effects of low-quality labels on gradient generation and enhance the competitiveness of high-quality anchor frames, a superior gradient gain allocation strategy is employed. Furthermore, the strategy introduces the Wise-IoU (WIoU), a dynamic non-monotonic focusing mechanism for bounding box loss, to further enhance the accuracy and speed of the object detection algorithm. The algorithm’s effectiveness is validated through experiments conducted on the 3L-TT100K traffic sign dataset, showcasing a mean average precision (mAP) of 94.1% in traffic sign detection. This mAP surpasses the performance of the YOLOv5(s) algorithm by 4.76% and outperforms the baseline algorithm. Additionally, the algorithm achieves a detection speed of 62.50 frames per second, which is much better than the baseline algorithm. Full article

In recent years, as more and more vehicles request service from roadside units (RSU), the vehicle-to-infrastructure (V2I) communication links are under tremendous pressure. This paper first proposes a dynamic dense traffic flow model under the condition of fading channel. Based on this, the reliability is redefined according to the real-time location information of vehicles. The on-board units (OBU) migrate intensive computing tasks to the appropriate RSU to optimize the execution time and calculating cost at the same time. In addition, competitive delay is introduced into the model of execution time, which can describe the channel resource contention and data conflict in dynamic scenes of the internet of vehicles (IoV). Next, the task scheduling for RSU is formulated as a multi-objective optimization problem. In order to solve the problem, a task scheduling algorithm based on a reliability constraint (TSARC) is proposed to select the optimal RSU for task transmission. When compared with the genetic algorithm (GA), there are some improvements of TSARC: first, the quick non-dominated sorting is applied to layer the population and reduce the complexity. Second, the elite strategy is introduced with an excellent nonlinear optimization ability, which ensures the diversity of optimal individuals and provides different preference choices for passengers. Third, the reference point mechanism is introduced to reserve the individuals that are non-dominated and close to reference points. TSARC’s Pareto based multi-objective optimization can comprehensively measure the overall state of the system and flexibly schedule system resources. Furthermore, it overcomes the defects of the GA method, such as the determination of the linear weight value, the non-uniformity of dimensions among objectives, and poor robustness. Finally, numerical simulation results based on the British Highway Traffic Flow Data Set show that the TSARC performs better scalability and efficiency than other methods with different numbers of tasks and traffic flow densities, which verifies the previous theoretical derivation. Full article

Surface defects of fiber-reinforced resin matrix composites (FRRMCs) adversely affect their appearance and performance. To accurately and efficiently detect the three-dimensional (3D) surface defects of FRRMCs, a novel lightweight and two-stage semantic segmentation network, i.e., Mask-Point, is proposed. Stage 1 of Mask-Point is the multi-head 3D region proposal extractors (RPEs), generating several 3D regions of interest (ROIs). Stage 2 is the 3D aggregation stage composed of the shared classifier, shared filter, and non-maximum suppression (NMS). The two stages work together to detect the surface defects. To evaluate the performance of Mask-Point, a new 3D surface defects dataset of FRRMCs containing about 120 million points is produced. Training and test experiments show that the accuracy and the mean intersection of union (mIoU) increase as the number of different 3D RPEs increases in Stage 1, but the inference speed becomes slower when the number of different 3D RPEs increases. The best accuracy, mIoU, and inference speed of the Mask-Point model could reach 0.9997, 0.9402, and 320,000 points/s, respectively. Moreover, comparison experiments also show that Mask-Point offers relatively the best segmentation performance compared with several other typical 3D semantic segmentation networks. The mIoU of Mask-Point is about 30% ahead of the sub-optimal 3D semantic segmentation network PointNet. In addition, a distributed surface defects detection system based on Mask-Point is developed. The system is applied to scan real FRRMC products and detect their surface defects, and it achieves the relatively best detection performance in competition with skilled human workers. The above experiments demonstrate that the proposed Mask-Point could accurately and efficiently detect 3D surface defects of FRRMCs, and the Mask-Point also provides a new potential solution for the 3D surface defects detection of other similar materials Full article

Most of the methods for real-time semantic segmentation do not take into account temporal information when working with video sequences. This is counter-intuitive in real-world scenarios where the main application of such methods is, precisely, being able to process frame sequences as quickly and accurately as possible. In this paper, we address this problem by exploiting the temporal information provided by previous frames of the video stream. Our method leverages a previous input frame as well as the previous output of the network to enhance the prediction accuracy of the current input frame. We develop a module that obtains feature maps rich in change information. Additionally, we incorporate the previous output of the network into all the decoder stages as a way of increasing the attention given to relevant features. Finally, to properly train and evaluate our methods, we introduce CityscapesVid, a dataset specifically designed to benchmark semantic video segmentation networks. Our proposed network, entitled FASSVid improves the mIoU accuracy performance over a standard non-sequential baseline model. Moreover, FASSVid obtains state-of-the-art inference speed and competitive mIoU results compared to other state-of-the-art lightweight networks, with significantly lower number of computations. Specifically, we obtain 71% of mIoU in our CityscapesVid dataset, running at 114.9 FPS on a single NVIDIA GTX 1080Ti and 31 FPS on the NVIDIA Jetson Nano embedded board with images of size $1024\times 2048$ and $512\times 1024$, respectively. Full article

Instance segmentation of sonar images is an effective method for underwater target recognition. However, the mismatch among positioning accuracy found by boxIoU and classification confidence, which is used as NMS score in current instance segmentation models; and the high annotation cost of sonar images, are two major problems in the task. To tackle these problems, in this paper, we present a novel instance segmentation method called Mask-Box Scoring R-CNN and embedded it in our proposed deep active learning framework. For the mismatch problem between boxIoU and NMS score, Mask-Box Scoring R-CNN uses a boxIoU head to predict the quality of the bounding boxes. We amend the non-maximum suppression (NMS) score predicted by BoxIoU to preserve high-quality bounding boxes in inference flow. To deal with the annotating problem, we propose a triplets-measure-based active learning (TBAL) method and a balanced-sampling method applicable for deep learning. The TBAL method evaluates the amount of information of unlabeled samples from the aspects of classification confidence, positioning accuracy, and mask quality. The balanced-sampling method selects hard samples from the dataset to train the model to improve performance. The experimental results show that Mask-Box Scoring R-CNN achieves improvements of 1% in boxAP and 1.3% boxAP on our sonar image dataset compared with Mask Scoring R-CNN and Mask R-CNN, respectively. The active learning framework with TBAL and balanced sampling can achieve a competitive performance with less labeled samples than other frameworks, which can better facilitate underwater target recognition. Full article

The blooming of internet of things (IoT) services calls for a paradigm shift in the design of communications systems. Short data packets sporadically transmitted by a multitude of low-cost low-power terminals require a radical change in relevant aspects of the protocol stack. For example, scheduling-based approaches may become inefficient at the medium access (MAC) layer, and alternatives such as uncoordinated access policies may be preferred. In this context random access (RA) in its simplest form, i.e., additive links on-line Hawaii area (ALOHA), may again become attractive as also proved by a number of technologies adopting it. The use of forward error correction (FEC) can improve its performance, yet a comprehensive analytical model including this aspect is still missing. In this paper, we provide a first attempt by deriving exact expressions for the packet loss rate and spectral efficiency of ALOHA with FEC, and extend the result also to time- and frequency-asynchronous ALOHA aided by FEC. We complement our study with extensive evaluations of the expressions for relevant cases of study, including an IoT system served by low-Earth orbit (LEO) satellites. Non-trivial outcomes show how time- and frequency-asynchronous ALOHA particularly benefit from the presence of FEC and become competitive with ALOHA. Full article

Extracting roads from remote sensing images can support a range of geo-information applications. However, it is challenging due to factors such as the complex distribution of ground objects and occlusion of buildings, trees, shadows, etc. Pixel-wise classification often fails to predict road connectivity and thus produces fragmented road segments. In this paper, we propose a multi-scale strip pooling network (MSPNet) to learn the linear features of roads. Motivated by the strip pooling being more aligned with the shape of roads, which are long-span and narrow, we develop a multi-scale strip pooling (MSP) module that utilizes strip pooling layers with long but narrow kernel shapes to capture multi-scale long-range context from horizontal and vertical directions. The proposed MSP module focuses on establishing relationships along the road region to guarantee the connectivity of roads. Considering the complex distribution of ground objects, the spatial pyramid pooling is applied to enhance the learning ability of complex features in different sub-regions. In addition, to alleviate the problem caused by an imbalanced distribution of road and non-road pixels, we use binary cross-entropy and dice-coefficient loss functions to jointly train our proposed deep learning model. Then, we perform ablation experiments to adjust the loss contributions to suit the task of road extraction. Comparative experiments on a popular benchmark DeepGlobe dataset demonstrate that our proposed MSPNet establishes new competitive results in both IoU and F1-score. Full article

This paper uses the non-competitive I-O model and the Tapio decoupling model to comprehensively analyze the decoupling relationship between the output of the product sector in China and its embodied carbon emissions under trade openness. For this purpose, the Chinese input and output data in 2002, 2005, 2007, 2010, and 2012 are used. This approach is beneficial to identify the direct mechanism for the increased carbon emission in China from a micro perspective and provides a new perspective for the subsequent study about low-carbon economy. The obtained empirical results are as follows: (1) From overall perspective, the decoupling elasticity between the output of the product sector and its embodied carbon emissions decreased. Output and embodied carbon emissions showed a growth link from 2002 to 2005 and a weak decoupling relationship for the rest of the study period. (2) Among the 28 industries in the product sector, the increased growth rate of output in more and more product sectors was no longer accompanied by large CO₂ emissions. The number of industries with strong decoupling relationships between output and embodied carbon emissions increased. (3) From the perspective of three industries, the output and embodied carbon emissions in the second and third industries exhibited a growth link only from 2002 to 2005; the three industries presented weak or strong decoupling for the rest of the study period. Through empirical analysis, this paper mainly through the construction of ecological and environmental protection of low carbon agriculture, low carbon cycle industrial system, as well as intensive and efficient service industry to reduce the carbon emissions of China’s product sector. Full article