Salient Target Research Articles

A weighted block cooperative sparse representation algorithm based on visual saliency dictionary

AbstractUnconstrained face images are interfered by many factors such as illumination, posture, expression, occlusion, age, accessories and so on, resulting in the randomness of the noise pollution implied in the original samples. In order to improve the sample quality, a weighted block cooperative sparse representation algorithm is proposed based on visual saliency dictionary. First, the algorithm uses the biological visual attention mechanism to quickly and accurately obtain the face salient target and constructs the visual salient dictionary. Then, a block cooperation framework is presented to perform sparse coding for different local structures of human face, and the weighted regular term is introduced in the sparse representation process to enhance the identification of information hidden in the coding coefficients. Finally, by synthesising the sparse representation results of all visual salient block dictionaries, the global coding residual is obtained and the class label is given. The experimental results on four databases, that is, AR, extended Yale B, LFW and PubFig, indicate that the combination of visual saliency dictionary, block cooperative sparse representation and weighted constraint coding can effectively enhance the accuracy of sparse representation of the samples to be tested and improve the performance of unconstrained face recognition.

Infrared and visible image fusion algorithm based on a cross-layer densely connected convolutional network.

To preserve the saliency of targets in infrared images and the textures in visible images, a novel infrared and visible image fusion method, to the best of our knowledge, is proposed. First, we design a densely connected convolutional network that contains an encoder, fusion, and decoder to minimize the omission of source image effective information. Then, a loss function based on the variational model is designed to retain the thermal radiation information of the infrared image and the details of the visible image to the greatest extent. The experimental results show that the proposed method outperforms state-of-the-art methods in terms of six metrics and better preserves the clear target and textures of infrared and visible images.

Research on defect detection algorithm of complex texture ceramic tiles based on visual attention mechanism

Aiming at the difficulty of detecting the surface defects of complex texture tiles, a salient target detection method based on the human visual attention mechanism is proposed and used for the detection of tile surface defects. Firstly, the image of ceramic tile surface is pretreated using the single-scale SSR light correction method and bilateral filtering method; Secondly, according to the principle of contrast and high-frequency suppression in the visual attention mechanism, aiming at the "imaging" and "aggregation" characteristics of complex background textures, a detection model based on the visual attention mechanism is established to determine and mark defects.According to the contrast principle and high-frequency suppression principle in visual attention mechanism, feature extraction of ceramic tile surface is carried out. Then, the image color patch weight salient map and image feature fused salient map are obtained, and the two maps are fused according to the image saliency criteria.Finally, the marked ceramic tile defects are determined and marked.Finally the marked ceramic tile defects are obtained. This defect detection algorithm and the other two algorithms are applied to three kinds of randomly selected complex texture ceramic tiles. The experimental results show that compared with other algorithms, our algorithm can achieve a comprehensive detection rate of more than 96% for complex texture ceramic tiles, and can obtain a good effect of ceramic tile defect detection as well.

PIAFusion: A progressive infrared and visible image fusion network based on illumination aware

Infrared and visible image fusion aims to synthesize a single fused image containing salient targets and abundant texture details even under extreme illumination conditions. However, existing image fusion algorithms fail to take the illumination factor into account in the modeling process. In this paper, we propose a progressive image fusion network based on illumination-aware, termed as PIAFusion, which adaptively maintains the intensity distribution of salient targets and preserves texture information in the background. Specifically, we design an illumination-aware sub-network to estimate the illumination distribution and calculate the illumination probability. Moreover, we utilize the illumination probability to construct an illumination-aware loss to guide the training of the fusion network. The cross-modality differential aware fusion module and halfway fusion strategy completely integrate common and complementary information under the constraint of illumination-aware loss. In addition, a new benchmark dataset for infrared and visible image fusion, i.e., Multi-Spectral Road Scenarios (available at https://github.com/Linfeng-Tang/MSRS), is released to support network training and comprehensive evaluation. Extensive experiments demonstrate the superiority of our method over state-of-the-art alternatives in terms of target maintenance and texture preservation. Particularly, our progressive fusion framework could round-the-clock integrate meaningful information from source images according to illumination conditions. Furthermore, the application to semantic segmentation demonstrates the potential of our PIAFusion for high-level vision tasks. Our codes will be available at https://github.com/Linfeng-Tang/PIAFusion.

Picking out the Impurities: Attention-based Push-Grasping in Dense Clutter

Abstract“Picking out the impurities” is a typical scenario in production line which is both time consuming and laborious. In this article, we propose a target-oriented robotic push-grasping system which is able to actively discover and pick the impurities in dense environments with the synergies between pushing and grasping actions. First, we propose an attention module, which includes target saliency detection and density-based occluded-region inference. Without the necessity of expensive labeling of semantic segmentation, our attention module can quickly locate the targets in the view or predict the candidate regions where the targets are most likely to be occluded. Second, we propose a push–grasp synergy framework to sequentially select proper actions in different situations until all targets are picked out. Moreover, we introduce an active pushing mechanism based on a novel metric, namely Target-Centric Dispersion Degree (TCDD) for better grasping. TCDD describes whether the targets are isolated from the surrounding objects. With this metric, the robot becomes more focused on the actions around the targets and push irrelevant objects away. Experimental results on both simulated environment and real-world environment show that our proposed system outperforms several baseline approaches,which also has the capability to be generalized to new scenarios.

Salient object detection network with center pooling and distance-weighted affinity loss function

Due to the rapid development of deep learning, salient target detection models have achieved significant progress. However, these models ignore the center prior in the collected images, that is, there is a greater probability of being in the center area for the salient object. The hidden information in the image is not fully utilized, which hinders further improvement of the detection result. We propose a center pooling algorithm that realizes the combination of image center prior and deep learning. The algorithm dynamically modifies the size of the receptive field according to the probability of the existence of salient objects and solves the problem of ignoring the importance of different regions when using the same pooling receptive field in the entire image by the standard pooling method. At the same time, the proportion of the foreground in the image is increased, and the problem of category imbalance is alleviated. In addition, a distance-weighted affinity loss function is also proposed to supplement the keyspace information by learning the grouping or separating force between pixel pairs at different distances and increase the confidence of edge regions. To further improve the detection results, a background detection subnetwork is introduced. Through the interactive learning of the foreground and background subnetworks, the opposite relationship between the predicted maps is explored. Finally, experiments on six different related datasets are conducted, and comparisons with other models prove that our model achieves the best performance.

Charting cognition: Mapping public understanding of COVID-19.

The COVID-19 pandemic of the last 2 years (and counting) disrupted commerce, travel, workplaces, habits, and—of course—health, the world over. This study aimed to capture snapshots of the perceptions and misperceptions of COVID-19 among 27 participants from three US municipalities. These perspectives are analyzed through thematic analyses and concept maps. Such snapshots, particularly as viewed through the lens of narrative sense-making theory, capture a sample of cognitions at this unique moment in history: a little over 1 year into the COVID-19 pandemic. Findings suggest that the (mis)perceptions captured are predominantly conveyed via narratives of participants’ personal experiences, and that the themes of attitudes toward precautionary measures, uncertainty, and the muddied science communication environment are prevalent. These themes suggest several salient targets for future research and current science communication, such as a focus on basic explainers, vaccinations’ safety and effectiveness and the necessity of uncertainty in the practice of science.

Urgency-Based color coding to support visual search in displays for supervisory control of multiple unmanned aircraft systems

This study derives and validates a color-coding concept for system states in icon-based displays for supervisory control of multiple unmanned aircraft systems. In aviation, safe indications are typically coded with green, caution indications with amber and warnings with red. This study proposes to adjust the salience of indications according to their intended urgency. For this purpose, a numerical measure for target salience was developed based on findings from the visual search literature. The measure uses differences in color space between target, distractor and background colors to estimate target salience. The measure is computed by dividing the color difference between target and distractor by the color difference between distractor and background and multiplying the result by the division of the target-distractor and distractor-background color differences. The resulting target salience value increases when decreasing the difference between distractor and background color as well as increasing differences between target and distractor and target and background colors. The results from applying the target salience measure suggest to code safe indications with grey instead of green, caution indications with yellow and warning and advisory indications with red and cyan, respectively. Two visual search experiments were conducted to validate the target salience measure and the proposed color-coding concept. Experiment one (n = 32) was a within-subject design comparing reaction time differences between yellow and amber targets among green and grey distractors for four set size conditions. Experiment two (n = 62) was a mixed-subject design comparing reaction time differences between red, yellow and cyan targets among grey and green distractors for four set size and 18 eccentricity conditions. The results show that target saliences diminish with increasing distractor salience, which could particularly be observed for yellow targets among green distractors. Red and yellow targets among less salient, grey distractors allow for a fast, parallel search. Further, the results show that cyan targets among grey distractors facilitate parallel search, although the reaction times are longer than for red and yellow targets. The findings support the proposed numerical measure for target salience and favor the color-coding concept thus suggesting to use grey for safe, and yellow and red for safety critical indications. Cyan is suited for coding advisory indications that are less critical. The results are transferable to the design of icon-based displays in similar domains, such as highly automated ground and maritime transport systems.

Dual-Fusion Active Contour Model with Semantic Information for Saliency Target Extraction of Underwater Images

Underwater vision research is the foundation of marine-related disciplines. The target contour extraction is significant for target tracking and visual information mining. Aiming to resolve the problem that conventional active contour models cannot effectively extract the contours of salient targets in underwater images, we propose a dual-fusion active contour model with semantic information. First, the saliency images are introduced as semantic information and salient target contours are extracted by fusing Chan–Vese and local binary fitting models. Then, the original underwater images are used to supplement the missing contour information by using the local image fitting. Compared with state-of-the-art contour extraction methods, our dual-fusion active contour model can effectively filter out background information and accurately extract salient target contours. Moreover, the proposed model achieves the best results in the quantitative comparison of MAE (mean absolute error), ER (error rate), and DR (detection rate) indicators and provides reliable prior knowledge for target tracking and visual information mining.

Application of the Multi-Process Action Control Model to Predict Physical Activity During Late Adolescence

The aim of this study was to apply the Multi-Process Action Control model to examine how the additions of regulatory and reflexive processes predict physical activity (PA) behaviors among adolescents. Our sample included 1,176 Grade 11 students (Mage = 15.85 ± 0.38) recruited from a large school board in Southern Ontario. Participants completed a questionnaire including measures of self-reported PA and PA cognitions derived from the Multi-Process Action Control model. Results found the reflective process explaining 16.5% of the variance in PA, with the additions of regulatory and reflexive processes significantly improving the explained variance by 5.1% and 8.2%, respectively. Final models revealed coping planning (estimate = 45.10, p = .047), identity (estimate = 55.82, p < .001), and habit (estimate = 64.07, p < .001) as significant predictors of PA. Findings reinforce the need for integrative models to better understand PA, with coping planning, habit formation, and development of an active identity to be salient targets for intervention during adolescence.

Significant target analysis and detail preserving based infrared and visible image fusion

Infrared and visible image fusion aims to integrate the prominent infrared target and visible texture details as much as possible. However, infrared images are susceptible to noise pollution during the transmission process, which could reduce the quality of the fused image. To solve this problem, a novel significant target analysis and detail preserving based scheme is proposed for infrared and visible image fusion tasks. Since the infrared image has high visual saliency, the noise pixels in the infrared image are often retained into the fused result by most traditional fusion methods. To better highlight the infrared targets while reducing interference from noise pixels, an infrared target detection model is proposed based on boundary-aware salient target detection network (BASNet) and significant target analysis. To fully preserve the visual details, we employ a weight mean curvature (WMC) based multiscale transform (MST) fusion scheme which can effectively suppress noise and preserve valuable details. Difference of Gaussian (DoG) is also applied to enhance the overall details in the proposed fusion scheme. Qualitative and quantitative experimental results demonstrate that the proposed method can generate fused images with abundant texture details and prominent infrared targets, which is superior to some of the existing methods in visual effect and objective quality evaluations.

An EEG study of the combined effects of top-down and bottom-up attentional selection under varying task difficulty.

We examined the effect of combined top-down and bottom-up attentional control sources, using known attention-related EEG components that are thought to reflect target selection (N2pc) and distractor suppression (PD ). We used endogenous cues (valid vs. neutral) for top-down attentional control, and salience in the form of color singletons (either the target or a distractor) for bottom-up attentional control in visual search. Crucially, in two experiments, the task was of increasing difficulty, reporting the orientation of a tilted target (Experiment 1), or the position of a small gap within the target among tilted non-targets (Experiment 2). Our results showed strong cueing effects on RT and accuracy in both experiments, demonstrating a general facilitation of responses to validly cued targets. Whereas the processing of salient targets was not improved compared with non-salient targets, the presence of a salient distractor consistently worsened performance. The N2pc and PD were only observed in trials where targets were preceded by neutral cues in Experiment 1, and for validly cued targets and salient neutrally cued targets in Experiment 2. A cueing effect was found on the PD in Experiment 1, showing an amplitude reduction in trials where the target was validly cued. These results support the idea that bottom-up attentional allocation occurs only when top-down allocation of attention is absent or inefficient. Therefore, these results indicate that attentional selection and suppression during visual search are both influenced by top-down cueing and give support to theories that focus on the interaction between the two types of attention.

Visual Saliency via Multiscale Analysis in Frequency Domain and Its Applications to Ship Detection in Optical Satellite Images.

This article proposes a bottom-up visual saliency model that uses the wavelet transform to conduct multiscale analysis and computation in the frequency domain. First, we compute the multiscale magnitude spectra by performing a wavelet transform to decompose the magnitude spectrum of the discrete cosine coefficients of an input image. Next, we obtain multiple saliency maps of different spatial scales through an inverse transformation from the frequency domain to the spatial domain, which utilizes the discrete cosine magnitude spectra after multiscale wavelet decomposition. Then, we employ an evaluation function to automatically select the two best multiscale saliency maps. A final saliency map is generated via an adaptive integration of the two selected multiscale saliency maps. The proposed model is fast, efficient, and can simultaneously detect salient regions or objects of different sizes. It outperforms state-of-the-art bottom-up saliency approaches in the experiments of psychophysical consistency, eye fixation prediction, and saliency detection for natural images. In addition, the proposed model is applied to automatic ship detection in optical satellite images. Ship detection tests on satellite data of visual optical spectrum not only demonstrate our saliency model's effectiveness in detecting small and large salient targets but also verify its robustness against various sea background disturbances.

STAug: Copy-Paste Based Image Augmentation Technique Using Salient Target

High-quality, large-capacity data are essential for training a deep learning vision model. However, to construct crop image data, absolute growth time is required for crop growth. In addition, it is characterized by unbalanced data, with fewer abnormal data than normal data. Therefore, building high-quality, large-scale datasets is challenging. Many studies have used data augmentation of plant images to solve this problem. However, plants require data augmentation that does not compromise their color, texture, or shape. This study proposes the use of salient target augmentation (STAug) as a data augmentation technique to protect the colors and shapes of plant images. The proposed method pastes one image’s salient target into a different image to mix the two images. It uses a salient object detection model to generate a salient object mask of the plant. Using the generated mask, a salient target was identified and cropped in a plant image, and the cropped image data were pasted to different background data for augmentation. Concat mask, a combination of each image’s salient object mask, was designed to create the label of the generated image. It is possible to create a rigid classification model by augmenting the data without damaging the plant features. To verify the performance of the proposed STAug, we compared its performance with that of other data-augmentation policies. When STAug and other augmentation techniques were applied in combination, an accuracy of 0.9733 was achieved. We demonstrated a better classification performance than when it was not applied.

Image fusion in the loop of high-level vision tasks: A semantic-aware real-time infrared and visible image fusion network

Infrared and visible image fusion aims to synthesize a single fused image that not only contains salient targets and abundant texture details but also facilitates high-level vision tasks. However, the existing fusion algorithms unilaterally focus on the visual quality and statistical metrics of fused images but ignore the demands of high-level vision tasks. To address these challenges, this paper bridges the gap between image fusion and high-level vision tasks and proposes a semantic-aware real-time image fusion network (SeAFusion). On the one hand, we cascade the image fusion module and semantic segmentation module and leverage the semantic loss to guide high-level semantic information to flow back to the image fusion module, which effectively boosts the performance of high-level vision tasks on fused images. On the other hand, we design a gradient residual dense block (GRDB) to enhance the description ability of the fusion network for fine-grained spatial details. Extensive comparative and generalization experiments demonstrate the superiority of our SeAFusion over state-of-the-art alternatives in terms of maintaining pixel intensity distribution and preserving texture detail. More importantly, the performance comparison of various fusion algorithms in task-driven evaluation reveals the natural advantages of our framework in facilitating high-level vision tasks. In addition, the superior running efficiency allows our algorithm to be effortlessly deployed as a real-time pre-processing module for high-level vision tasks. The source code will be released at https://github.com/Linfeng-Tang/SeAFusion.

Salient Targets and Fear of Falling Changed the Gait Pattern and Joint Kinematic of Older Adults

Salient Targets and Fear of Falling Changed the Gait Pattern and Joint Kinematic of Older Adults

Glucuronoxylomannan: the salient polysaccharide in cryptococcal immunity

&lt;abstract&gt; &lt;p&gt;Cryptococcal meningitis (CM) is a dominant cause of morbidity and mortality among patients with human immunodeficiency virus/ acquired immune deficiency syndrome (HIV/AIDS) caused by &lt;italic&gt;Cryptococcus neoformans&lt;/italic&gt; and &lt;italic&gt;Cryptococcus gattii&lt;/italic&gt; species complex. The complex is composed of closely related members, yet with diverse epidemiology, pathogenesis, and drug-resistant pattern. Cell-mediated immunity is the strongest pillar in immunity to cryptococcosis, further worsening HIV/AIDS patients' scenario. Antifungal resistance and immune evasion again tilt the host-parasite balance in favor of the fungal pathogen. In this regard, researchers are actively challenged to discover immunotherapy and vaccine for CM, to produce specific treatment and prevention that will address CM conventional therapeutics failure. As the major capsular polysaccharide of the &lt;italic&gt;Cryptococcus&lt;/italic&gt;, which is tightly linked to pathogenicity, immunogenicity, and immune evasion, the glucuronoxylomannan (GXM) is cardinally targeted for vaccine and immunotherapy development. Further, the amount of GXM shed in body fluids correlates with the disease severity. Herein, we reviewed the literature with the journey so far in line with GXM as the salient immunological target on cryptococcosis.&lt;/p&gt; &lt;/abstract&gt;

A Robust Space Target Detection Algorithm Based on Target Characteristics

In space-based observations, there are large numbers of targets with large size ranges and intensities in a given image. However, the current detection studies mostly emphasize the saliency of targets but neglect their distribution characteristics, which may result in incorrect or missed detection results. In this letter, an effective detection algorithm based on target characteristics is proposed to pursue good multitarget detection performance. First, a space target model is studied to obtain its characteristics. Next, operators are applied to reduce the influence and search for suspected target-center points. Then, local regions around these points are extracted and normalized to reduce the intensity differences among targets and the influence of light conditions. Finally, three features are designed to confirm these suspected regions and their scale information. Furthermore, the threshold setting process is intuitive and significant while keeping the developed method robust in different cases. Compared with other current algorithms, the proposed algorithm achieves superior detection performance in terms of the number of detected targets, accuracy rate, false alarm rate, and computational cost.

Adaptive Infrared and Visible Image Fusion Based on Visual Saliency and Hierarchical Bayesian

An adaptive infrared and visible image fusion method based on visual saliency and hierarchical Bayesian (AVSHB) which preserves the highest similarity between fused images and source images is proposed in this paper. Firstly, an effective salient edge preserving filter named SEPF is developed to decompose each source image into a base layer and a detail layer. A ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> norm gradient minimization is firstly derived and then embedded into a two-scale acceleration scheme in SEPF. Benefiting from the SEPF, the edges of salient regions can be preserved without distortion. Then, an adaptive fusion scheme is proposed which fully considers the characteristics of each layer. More concretely, we design a two-scale fusion strategy based on a visual saliency map (VSM) for the base layers, and a hierarchical Bayesian fusion model is derived for the detail layers. The experimental results on the TNO and RoadScene datasets and Nato_camp image sequence demonstrate that AVSHB favorably outperforms 16 related state-of-the-art fusion methods qualitatively and quantitatively. AVSHB can generate improved fusion results with sufficiently retaining salient targets and rich details from the source images.

Improved Flame Detection Algorithm Based on Salient Target Detection

Improved Flame Detection Algorithm Based on Salient Target Detection