Irrelevant Regions Research Articles

CrackNet: A Hybrid Model for Crack Segmentation with Dynamic Loss Function

Cracks are a common form of damage in infrastructure, posing significant risks to both personal safety and property. Along with the development of deep learning, visual-based crack automatic detection has been widely studied. However, this task is still challenging due to complex crack topology, noisy backgrounds, unbalanced categories, etc. To address these challenges, this research proposes a novel hybrid network, named CrackNet, which leverages the strengths of both CNN and transformer. On the encoder side, CNNs are employed to extract multi-level local features, while transformers are used to model global dependencies. Additionally, a strip pooling module is introduced to suppress irrelevant regions and enhance the network’s ability to segment narrow and elongated cracks. On the decoder side, an attention-based skip connection strategy and a mixed up-sampling module are implemented to restore detailed information. Furthermore, a joint learning loss combining Dice and cross-entropy with dynamic weighting is proposed to mitigate the effects of severe class imbalance. CrackNet is trained and evaluated on three public crack datasets, and experimental results show that the proposed model outperforms several well-known deep neural networks, with a particularly noticeable improvement in recall rate.

Bilevel progressive homography estimation via correlative region-focused transformer

We propose a novel correlative region-focused transformer for accurate homography estimation by a bilevel progressive architecture. Existing methods typically consider the entire image features to establish correlations for a pair of input images, but irrelevant regions often introduce mismatches and outliers. In contrast, our network effectively mitigates the negative impact of irrelevant regions through a bilevel progressive homography estimation architecture. Specifically, in the outer iteration, we progressively estimate the homography matrix at different feature scales; in the inner iteration, we dynamically extract correlative regions and progressively focus on their corresponding features from both inputs. Moreover, we develop a quadtree attention mechanism based on the transformer to explicitly capture the correspondence between the input images, localizing and cropping the correlative regions for the next iteration. This progressive training strategy enhances feature consistency and enables precise alignment with comparable inference rates. Extensive experiments on qualitative and quantitative comparisons show that the proposed method exhibits competitive alignment results while reducing the mean average corner error (MACE) on the MS-COCO dataset compared to previous methods, without increasing additional parameter cost.

A Patch-Level Region-Aware Module with a Multi-Label Framework for Remote Sensing Image Captioning

Recent Transformer-based works can generate high-quality captions for remote sensing images (RSIs). However, these methods generally feed global or grid visual features to a Transformer-based captioning model for associating cross-modal information, which limits performance. In this work, we investigate unexplored ideas for a remote sensing image captioning task, using a novel patch-level region-aware module with a multi-label framework. Due to an overhead perspective and a significantly larger scale in RSIs, a patch-level region-aware module is designed to filter the redundant information in the RSI scene, which benefits the Transformer-based decoder by attaining improved image perception. Technically, the trainable multi-label classifier capitalizes on semantic features as supplementary to the region-aware features. Moreover, modeling the inner relations of inputs is essential for understanding the RSI. Thus, we introduce region-oriented attention, which associates region features and semantic labels, omits the irrelevant regions to highlight relevant regions, and learns related semantic information. Extensive qualitative and quantitative experimental results show the superiority of our approach on the RSICD, UCM-Captions, and Sydney-Captions. The code for our method will be publicly available.

Polyp-Mamba: A Hybrid Multi-Frequency Perception Gated Selection Network for polyp segmentation

Accurate segmentation of polyps in the colorectal region is crucial for medical diagnosis and the localization of polyp areas. However, challenges arise from blurred boundaries due to the similarity between polyp edges and surrounding tissues, variable polyp morphology, and speckle noise. To address these challenges, we propose a Hybrid Multi-Frequency Perception Gated Selection Network (Polyp-Mamba) for precise polyp segmentation. First, we design a dual multi-frequency fusion encoder that employs Mamba and ResNet to quickly and effectively learn global and local features in polyp images. Specifically, we incorporate a novel Hybrid Multi-Frequency Fusion Module (HMFM) within the encoder, using discrete cosine transform to analyze features from multiple spectral perspectives. This approach mitigates the issue of blurred polyp boundaries caused by their similarity to surrounding tissues, effectively integrating local and global features. Additionally, we construct a Gated Selection Decoder to suppress irrelevant feature regions in the encoder and introduce deep supervision to guide decoder features to align closely with the labels. We conduct extensive experiments using five commonly used polyp test datasets. Comparisons with 14 state-of-the-art segmentation methods demonstrate that our approach surpasses traditional methods in sensitivity to different polyp images, robustness to variations in polyp size and shape, speckle noise, and distribution similarity between surrounding tissues and polyps. Overall, our method achieves superior mDice scores on five polyp test datasets compared to state-of-the-art methods, indicating better performance in polyp segmentation.

A unified feature-motion consistency framework for robust image matching

Establishing reliable feature matches between a pair of images in various scenarios is a long-standing open problem in photogrammetry. Attention-based detector-free matching with coarse-to-fine architecture has been a typical pipeline to build matches, but the cross-attention module with global receptive field may compromise the structural local consistency by introducing irrelevant regions (outliers). Motion field can maintain structural local consistency under the assumption that matches for adjacent features should be spatially proximate. However, motion field can only estimate local displacements between consecutive images and struggle with long-range displacements estimation in large-scale variation scenarios without spatial correlation priors. Moreover, large-scale variations may also disrupt the geometric consistency with the application of mutual nearest neighbor criterion in patch-level matching, making it difficult to recover accurate matches. In this paper, we propose a unified feature-motion consistency framework for robust image matching (MOMA), to maintain structural consistency at both global and local granularity in scale-discrepancy scenarios. MOMA devises a motion consistency-guided dependency range strategy (MDR) in cross attention, aggregating highly relevant regions within the motion consensus-restricted neighborhood to favor true matchable regions. Meanwhile, a unified framework with hierarchical attention structure is established to couple local motion field with global feature correspondence. The motion field provides local consistency constraints in feature aggregation, while feature correspondence provides spatial context prior to improve motion field estimation. To alleviate geometric inconsistency caused by hard nearest neighbor criterion, we propose an adaptive neighbor search (soft) strategy to address scale discrepancy. Extensive experiments on three datasets demonstrate that our method outperforms solid baselines, with AUC improvements of 4.73/4.02/3.34 in two-view pose estimation task at thresholds of 5°/10°/20° on Megadepth test, and 5.94% increase of accuracy at threshold of 1px in homography task on HPatches datasets. Furthermore, in the downstream tasks such as 3D mapping, the 3D models reconstructed using our method on the self-collected SYSU UAV datasets exhibit significant improvement in structural completeness and detail richness, manifesting its high applicability in wide downstream tasks. The code is publicly available at https://github.com/BunnyanChou/MOMA.

CRENet: Crowd region enhancement network for multi-person 3D pose estimation

Recovering multi-person 3D poses from a single image is a challenging problem due to inherent depth ambiguities, including root-relative depth and absolute root depth. Current bottom-up methods show promising potential to mitigate absolute root depth ambiguity through explicitly aggregating global contextual cues. However, these methods treat the entire image region equally during root depth regression, ignoring the negative impact of irrelevant regions. Moreover, they learn shared features for both depths, each of which focuses on different information. This sharing mechanism may result in negative transfer, thus diminishing root depth prediction accuracy. To address these challenges, we present a novel bottom-up method, Crowd Region Enhancement Network (CRENet), incorporating a Feature Decoupling Module (FDM) and a Global Attention Module (GAM). FDM explicitly learns the discriminative feature for each depth through adaptively recalibrating its channel-wise responses and fusing multi-level features, which makes the model focus on each depth prediction separately and thus avoids the adverse effect of negative transfer. GAM highlights crowd regions while suppressing irrelevant regions using the attention mechanism and further refines the attention regions based on the confidence measure about the attention, which is beneficial to learn depth-related cues from informative crowd regions and facilitate root depth estimation. Comprehensive experiments on benchmarks MuPoTS-3D and CMU Panoptic demonstrate that our method outperforms the state-of-the-art bottom-up methods in absolute 3D pose estimation and is applicable to in-the-wild images, which also indicates that learning depth-specific features and suppressing the noise signals can significantly benefit multi-person absolute 3D pose estimation.

Application of U-Net Network Utilizing Multiattention Gate for MRI Segmentation of Brain Tumors.

Studies have shown that the type of low-grade glioma is associated with its shape. The traditional diagnostic method involves extraction of the tumor shape from MRIs and diagnosing the type of glioma based on corresponding relationship between the glioma shape and type. This method is affected by the MRI background, tumor pixel size, and doctors' professional level, leading to misdiagnoses and missed diagnoses. With the help of deep learning algorithms, the shape of a glioma can be automatically segmented, thereby assisting doctors to focus more on the diagnosis of glioma and improving diagnostic efficiency. The segmentation of glioma MRIs using traditional deep learning algorithms exhibits limited accuracy, thereby impeding the effectiveness of assisting doctors in the diagnosis. The primary objective of our research is to facilitate the segmentation of low-grade glioma MRIs for medical practitioners through the utilization of deep learning algorithms. In this study, a UNet glioma segmentation network that incorporates multiattention gates was proposed to address this limitation. The UNet-based algorithm in the coding part integrated the attention gate into the hierarchical structure of the network to suppress the features of irrelevant regions and reduce the feature redundancy. In the decoding part, by adding attention gates in the fusion process of low- and high-level features, important feature information was highlighted, model parameters were reduced, and model sensitivity and accuracy were improved. The network model performed image segmentation on the glioma MRI dataset, and the accuracy and average intersection ratio (mIoU) of the algorithm segmentation reached 99.7%, 87.3%, 99.7%, and 87.6%. Compared with the UNet, PSPNet, and Attention UNet network models, this network model has obvious advantages in accuracy, mIoU, and loss convergence. It can serve as a standard for assisting doctors in diagnosis.

Decoupled Cross-Modal Transformer for Referring Video Object Segmentation.

Referring video object segmentation (R-VOS) is a fundamental vision-language task which aims to segment the target referred by language expression in all video frames. Existing query-based R-VOS methods have conducted in-depth exploration of the interaction and alignment between visual and linguistic features but fail to transfer the information of the two modalities to the query vector with balanced intensities. Furthermore, most of the traditional approaches suffer from severe information loss in the process of multi-scale feature fusion, resulting in inaccurate segmentation. In this paper, we propose DCT, an end-to-end decoupled cross-modal transformer for referring video object segmentation, to better utilize multi-modal and multi-scale information. Specifically, we first design a Language-Guided Visual Enhancement Module (LGVE) to transmit discriminative linguistic information to visual features of all levels, performing an initial filtering of irrelevant background regions. Then, we propose a decoupled transformer decoder, using a set of object queries to gather entity-related information from both visual and linguistic features independently, mitigating the attention bias caused by feature size differences. Finally, the Cross-layer Feature Pyramid Network (CFPN) is introduced to preserve more visual details by establishing direct cross-layer communication. Extensive experiments have been carried out on A2D-Sentences, JHMDB-Sentences and Ref-Youtube-VOS. The results show that DCT achieves competitive segmentation accuracy compared with the state-of-the-art methods.

MRSAPose: Multi-level routing sparse attention for multi-person pose estimation

Multi-person Pose Estimation (MPPE) aims to reconstruct human poses by locating and connecting keypoints of individuals in input images. The variability of human poses and the complexity of scenes make MPPE reliant on both local details and global structures, and the absence of either can lead to the generation of deformed poses. With the emergence of Transformer, the performance of MPPE has been significantly improved. However, due to self-attention computing attention scores between each pair of positions, the current Transformer-based MPPE exhibits high quadratic complexity. To address these issues, this paper proposes a novel pose estimation model, MRSAPose. MRSAPose utilizes Multi-level Routing Sparse Attention (MRSA) to dynamically select relevant regions for attention, reducing computational complexity and mitigating the impact of irrelevant regions. Furthermore, MRSAPose constructs a Transformer-CNN Parallel Interaction Block (T-CP block) through MRSA and Recursive Residual Gated Convolution (Res-gnConv), facilitating parallel learning of global and local information. By relying on multi-level routing algorithms and high-order spatial interactions conducted by recursive processing of adjacent features, T-CP block helps MRSAPose effectively alleviates the issues of occlusion and misalignment in pose estimation. On multiple challenging keypoint datasets, MRSAPose outperforms current state-of-the-art algorithms, particularly excelling in crowded and occluded scenes.

Guiding attention in flow-based conceptual models through consistent flow and pattern visibility

A critical part of flow-based conceptual modeling, such as process modeling, is visualizing the logical and temporal sequence in which activities in a process should be completed. While there are established standards and recommendations, there is limited empirical research examining the influence of process model layout on model comprehension. To address this research gap, we conducted a controlled eye-tracking experiment with 70 participants comparing different layouts. The experimental results confirm that the visibility of control flow patterns is critical for assisting users with visual processing, particularly attentional allocation, when comprehending process models for both local comprehension tasks and tasks requiring cognitive integration of model components. In models with more directional changes, users’ visual attention is more drawn to irrelevant regions, but comprehension is less affected as long as patterns remain visible. Our findings not only elucidate how cognitive fit between a visual representation and a task can manifest itself and the perceptual benefits it brings, but they can also guide the automated layout of models in tools and complement practical process modeling guidelines.

Deep attention network for identifying ligand-protein binding sites

One of the critical aspects of structure-based drug design is to choose important druggable binding sites in the protein's crystallography structures. As experimental processes are costly and time-consuming, computational drug design using machine learning algorithms is recommended. Over recent years, deep learning methods have been utilized in a wide variety of research applications such as binding site prediction. In this study, a new combination of attention blocks in the 3D U-Net model based on semantic segmentation methods is used to improve localization of pocket prediction. The attention blocks are tuned to find which point and channel of features should be emphasized along spatial and channel axes. Our model's performance is evaluated through extensive experiments on several datasets from different sources, and the results are compared to the most recent deep learning-based models. The results indicate the proposed attention model (Att-UNet) can predict binding sites accurately, i.e. the overlap of the predicted pocket using the proposed method with the true binding site shows statistically significant improvement when compared to other state-of-the-art models. The attention blocks may help the model focus on the target structure by suppressing features in irrelevant regions.

Identification of Pepper Leaf Diseases Based on TPSAO-AMWNet.

Pepper is a high-economic-value agricultural crop that faces diverse disease challenges such as blight and anthracnose. These diseases not only reduce the yield of pepper but, in severe cases, can also cause significant economic losses and threaten food security. The timely and accurate identification of pepper diseases is crucial. Image recognition technology plays a key role in this aspect by automating and efficiently identifying pepper diseases, helping agricultural workers to adopt and implement effective control strategies, alleviating the impact of diseases, and being of great importance for improving agricultural production efficiency and promoting sustainable agricultural development. In response to issues such as edge-blurring and the extraction of minute features in pepper disease image recognition, as well as the difficulty in determining the optimal learning rate during the training process of traditional pepper disease identification networks, a new pepper disease recognition model based on the TPSAO-AMWNet is proposed. First, an Adaptive Residual Pyramid Convolution (ARPC) structure combined with a Squeeze-and-Excitation (SE) module is proposed to solve the problem of edge-blurring by utilizing adaptivity and channel attention; secondly, to address the issue of micro-feature extraction, Minor Triplet Disease Focus Attention (MTDFA) is proposed to enhance the capture of local details of pepper leaf disease features while maintaining attention to global features, reducing interference from irrelevant regions; then, a mixed loss function combining Weighted Focal Loss and L2 regularization (WfrLoss) is introduced to refine the learning strategy during dataset processing, enhancing the model's performance and generalization capabilities while preventing overfitting. Subsequently, to tackle the challenge of determining the optimal learning rate, the tent particle snow ablation optimizer (TPSAO) is developed to accurately identify the most effective learning rate. The TPSAO-AMWNet model, trained on our custom datasets, is evaluated against other existing methods. The model attains an average accuracy of 93.52% and an F1 score of 93.15%, demonstrating robust effectiveness and practicality in classifying pepper diseases. These results also offer valuable insights for disease detection in various other crops.

DSTAN: A Deformable Spatial-temporal Attention Network with Bidirectional Sequence Feature Refinement for Speckle Noise Removal in Thyroid Ultrasound Video.

Thyroid ultrasound video provides significant value for thyroid diseases diagnosis, but the ultrasound imaging process is often affected by the speckle noise, resulting in poor quality of the ultrasound video. Numerous video denoising methods have been proposed to remove noise while preserving texture details. However, existing methods still suffer from the following problems: (1) relevant temporal features in the low-contrast ultrasound video cannot be accurately aligned and effectively aggregated by simple optical flow or motion estimation, resulting in the artifacts and motion blur in the video; (2) fixed receptive field in spatial features integration lacks the flexibility of aggregating features in the global region of interest and is susceptible to interference from irrelevant noisy regions. In this work, we propose a deformable spatial-temporal attention denoising network to remove speckle noise in thyroid ultrasound video. The entire network follows the bidirectional feature propagation mechanism to efficiently exploit the spatial-temporal information of the whole video sequence. In this process, two modules are proposed to address the above problems: (1) a deformable temporal attention module (DTAM) is designed after optical flow pre-alignment to further capture and aggregate relevant temporal features according to the learned offsets between frames, so that inter-frame information can be better exploited even with the imprecise flow estimation under the low contrast of ultrasound video; (2) a deformable spatial attention module (DSAM) is proposed to flexibly integrate spatial features in the global region of interest through the learned intra-frame offsets, so that irrelevant noisy information can be ignored and essential information can be precisely exploited. Finally, all these refined features are rectified and merged through residual convolution blocks to recover the clean video frames. Experimental results on our thyroid ultrasound video (US-V) dataset and the DDTI dataset demonstrate that our proposed method exceeds 1.2 1.3dB on PSNR and has clearer texture detail compared to other state-of-the-art methods. In the meantime, the proposed model can also assist thyroid nodule segmentation methods to achieve more accurate segmentation effect, which provides an important basis for thyroid diagnosis. In the future, the proposed model can be improved and extended to other medical image sequence datasets, including CT and MRI slice denoising. The code and datasets are provided at https://github.com/Meta-MJ/DSTAN .

Plane-wave medical image reconstruction based on dynamic Criss-Cross attention and multi-scale convolution.

Plane-wave imaging is widely employed in medical imaging due to its ultra-fast imaging speed. However, the image quality is compromised. Existing techniques to enhance image quality tend to sacrifice the imaging frame rate. The study aims to reconstruct high-quality plane-wave images while maintaining the imaging frame rate. The proposed method utilizes a U-Net-based generator incorporating a multi-scale convolution module in the encoder to extract information at different levels. Additionally, a Dynamic Criss-Cross Attention (DCCA) mechanism is proposed in the decoder of the U-Net-based generator to extract both local and global features of plane-wave images while avoiding interference caused by irrelevant regions. In the reconstruction of point targets, the experimental images achieved a reduction in Full Width at Half Maximum (FWHM) of 0.0499 mm, compared to the Coherent Plane-Wave Compounding (CPWC) method using 75-beam plane waves. For the reconstruction of cyst targets, the simulated image achieved a 3.78% improvement in Contrast Ratio (CR) compared to CPWC. The proposed model effectively addresses the issue of unclear lesion sites in plane-wave images.

Segmentation of ethnic clothing patterns with fusion of multiple attention mechanisms

To address the challenges posed by diverse pattern-background elements, intricate details, and complex textures in the semantic segmentation of ethnic clothing patterns, this research introduces a novel semantic segmentation network model called MST-Unet (Mixed Swin Transformer U-net). The proposed model combines a U-shaped network structure with multiple attention mechanisms. The upper layers of the model employ classical convolutional operations, focusing on local relationships in the initial layers containing high-resolution details. In deeper layers, Swin Transformer modules are utilized, capable of efficient feature extraction with smaller spatial dimensions, maintaining performance while reducing computational burden. An attention gate mechanism is integrated into the decoder, contributing to enhanced performance in ethnic clothing pattern segmentation tasks by allowing the model to better capture crucial image features and achieve precise segmentation results. In visual comparisons of segmentation results, our proposed model demonstrates superior performance. The segmentation results exhibit more complete preservation of edge contours and fewer misclassifications in irrelevant regions within the images. In qualitative and quantitative experiments conducted on the ethnic clothing pattern dataset, our model achieves the highest Dice score for segmentation results in all four subclasses of ethnic clothing patterns. The average Dice score of our model reaches an impressive 89.80%, surpassing other algorithms in the same category. When compared to Deeplab_V3+, ResUnet, SwinUnet, and Unet networks, our model outperforms them by 7.72%, 5.09%, 5.05%, and 0.67%.

LSKANet: Long Strip Kernel Attention Network for Robotic Surgical Scene Segmentation.

Surgical scene segmentation is a critical task in Robotic-assisted surgery. However, the complexity of the surgical scene, which mainly includes local feature similarity (e.g., between different anatomical tissues), intraoperative complex artifacts, and indistinguishable boundaries, poses significant challenges to accurate segmentation. To tackle these problems, we propose the Long Strip Kernel Attention network (LSKANet), including two well-designed modules named Dual-block Large Kernel Attention module (DLKA) and Multiscale Affinity Feature Fusion module (MAFF), which can implement precise segmentation of surgical images. Specifically, by introducing strip convolutions with different topologies (cascaded and parallel) in two blocks and a large kernel design, DLKA can make full use of region- and strip-like surgical features and extract both visual and structural information to reduce the false segmentation caused by local feature similarity. In MAFF, affinity matrices calculated from multiscale feature maps are applied as feature fusion weights, which helps to address the interference of artifacts by suppressing the activations of irrelevant regions. Besides, the hybrid loss with Boundary Guided Head (BGH) is proposed to help the network segment indistinguishable boundaries effectively. We evaluate the proposed LSKANet on three datasets with different surgical scenes. The experimental results show that our method achieves new state-of-the-art results on all three datasets with improvements of 2.6%, 1.4%, and 3.4% mIoU, respectively. Furthermore, our method is compatible with different backbones and can significantly increase their segmentation accuracy. Code is available at https://github.com/YubinHan73/LSKANet.

Multi-scale features and attention guided for brain tumor segmentation

Brain tumor segmentation supplies a credible reference for clinical treatment and pathological research and facilitates practitioners to diagnose more accurately. However, since the randomness and complexity of tumor shape and location, automatic brain tumor segmentation remains an extremely challenging assignment. In this study, we build an end-to-end convolutional neural network with a U-shaped structure to implement the segmentation of three lesion regions. We propose a multi-scale context block and an attention guidance block to focus on the spatial information at different scales and the interdependence between feature channels to enhance network representations and boost the learning capability of the model. Specifically, the multi-scale context block draws rich feature information through 3D dilated convolution. The attention guidance block reduces the impact of learned redundant features and eliminates the interference of irrelevant regions in the overall global information. Our recommended approach is evaluated on the brain tumor segmentation 2020 validation data. The Dice scores of the enhancing tumor (ET), whole tumor (WT), and tumor core (TC) are 78.19%, 90.10%, and 83.98%, respectively. In addition, the practice is also carried out in 2019 online validation data, and the Dice scores of ET, WT, and TC are 77.31%, 89.64%, and 82.55%, respectively. Experimental results reveal that the recommended approach gains favorable performance in comparison with representative brain tumor segmentation approaches. Our present study would accurately and efficaciously segment the three brain lesion regions and has clinical practice value.

TCNet: Continuous Sign Language Recognition from Trajectories and Correlated Regions

A key challenge in continuous sign language recognition (CSLR) is to efficiently capture long-range spatial interactions over time from the video input. To address this challenge, we propose TCNet, a hybrid network that effectively models spatio-temporal information from Trajectories and Correlated regions. TCNet's trajectory module transforms frames into aligned trajectories composed of continuous visual tokens. This facilitates extracting region trajectory patterns. In addition, for a query token, self-attention is learned along the trajectory. As such, our network can also focus on fine-grained spatio-temporal patterns, such as finger movement, of a region in motion. TCNet's correlation module utilizes a novel dynamic attention mechanism that filters out irrelevant frame regions. Additionally, it assigns dynamic key-value tokens from correlated regions to each query. Both innovations significantly reduce the computation cost and memory. We perform experiments on four large-scale datasets: PHOENIX14, PHOENIX14-T, CSL, and CSL-Daily. Our results demonstrate that TCNet consistently achieves state-of-the-art performance. For example, we improve over the previous state-of-the-art by 1.5\% and 1.0\% word error rate on PHOENIX14 and PHOENIX14-T, respectively. Code is available at https://github.com/hotfinda/TCNet

Determination of quality classes for material extrusion additive manufacturing using image processing

Tools for implementing a systematic quality management are necessary for the use of material extrusion as an additive manufacturing process for products with high quality requirements. Well-defined quality classes are crucial for ensuring that the requirements for a product can be communicated transparently and that the existing properties can be evaluated. Furthermore, there is a lack of capable measurement equipment for the acquisition of process data during the production process. To address these challenges, the present paper introduces an image processing system that determines quality indicators for individual layers in terms of imperfect surface percentages and the number of imperfections. The central element of the hardware is an adaptive darkfield illumination, which leads to high-contrast images. In addition, five types of layer subareas are identified in a segmentation step. Unsupervised machine learning methods are then used to detect imperfections in each layer subarea. In the segmentation, the current layer can be distinguished from irrelevant image background regions with an F-measure of 0.981. For the layer-wise measurement of the quality indicators, relative measurement errors with standard deviations of 25 to 76.1% are found. After evaluating the capabilities of the image processing system, a proposal for limits of quality classes is derived by monitoring several material extrusion processes. For this purpose, three quality classes for each of the five layer subareas are deduced from the process scatter measured by the image processing system. The results are an important contribution to the industrialization of material extrusion in safety–critical areas such as medical technology or the aerospace industry.

Fast and high-precision compressible flowfield inference method of transonic airfoils based on attention UNet

Traditional numerical simulation methods for airfoil flowfields are complex and time-consuming, and deep learning-based inference methods for Reynolds-averaged Navier–Stokes equations (RANS) solutions of transonic airfoils have limitations in terms of their robustness and generalization. A novel data-driven inference method named as attention UNet (AU)-RANS is proposed for efficient and accurate prediction of flowfields around airfoils with strong compressibility and large-scale turbulent separation. First, to enhance the learning the boundary flow information and inference of the entire flowfield solution, an innovative data preprocessing method is proposed to convert the physical quantities and coordinate information of RANS solutions into neural network spatial information. Second, an attention mechanism is introduced in UNet to suppress feature responses in irrelevant background regions and enhance sensitivity to the geometrical features of the input airfoil and varying inflow conditions. The quantitative and qualitative analyses of AU-RANS inference results demonstrate that the well-trained model can effectively infer RANS solutions for airfoil flowfield and can accurately predict the shock waves and flow separation phenomena under high Mach number conditions with a large angle of attack.