Non-salient Objects Research Articles

Salient Objects in Clutter.

In this paper, we identify and address a serious design bias of existing salient object detection (SOD) datasets, which unrealistically assume that each image should contain at least one clear and uncluttered salient object. This design bias has led to a saturation in performance for state-of-the-art SOD models when evaluated on existing datasets. However, these models are still far from satisfactory when applied to real-world scenes. Based on our analyses, we propose a new high-quality dataset and update the previous saliency benchmark. Specifically, our dataset, called Salient Objects in Clutter (SOC), includes images with both salient and non-salient objects from several common object categories. In addition to object category annotations, each salient image is accompanied by attributes that reflect common challenges in common scenes, which can help provide deeper insight into the SOD problem. Further, with a given saliency encoder, e.g., the backbone network, existing saliency models are designed to achieve mapping from the training image set to the training ground-truth set. We therefore argue that improving the dataset can yield higher performance gains than focusing only on the decoder design. With this in mind, we investigate several dataset-enhancement strategies, including label smoothing to implicitly emphasize salient boundaries, random image augmentation to adapt saliency models to various scenarios, and self-supervised learning as a regularization strategy to learn from small datasets. Our extensive results demonstrate the effectiveness of these tricks. We also provide a comprehensive benchmark for SOD, which can be found in our repository: https://github.com/DengPingFan/SODBenchmark.

Deep Hypersphere Feature Regularization for Weakly Supervised RGB-D Salient Object Detection.

We propose a weakly supervised approach for salient object detection from multi-modal RGB-D data. Our approach only relies on labels from scribbles, which are much easier to annotate, compared with dense labels used in conventional fully supervised setting. In contrast to existing methods that employ supervision signals on the output space, our design regularizes the intermediate latent space to enhance discrimination between salient and non-salient objects. We further introduce a contour detection branch to implicitly constrain the semantic boundaries and achieve precise edges of detected salient objects. To enhance the long-range dependencies among local features, we introduce a Cross-Padding Attention Block (CPAB). Extensive experiments on seven benchmark datasets demonstrate that our method not only outperforms existing weakly supervised methods, but is also on par with several fully-supervised state-of-the-art models. Code is available at https://github.com/leolyj/DHFR-SOD.

Neural network based image object detection and tracking for security and surveillance

The saliency map obtained from the source image determines the efficacy of the traditional seam carving process. The importance map proposed in this paper is used to highlight the shadows and important objects in the images. It combines the saliency map, shadow map, and gradient map acquired from the image to discover the image’s prominent regions. The proposed map, when compared to others, highlights more distinct details with the state-of-the-art. The improved seam carving technique is paired with cropping and warping image retargeting operators in the suggested hybrid sequence. By labelling a picture with a class label and object localisation, the coordinates of the objects are generated using R-CNN object detection techniques. This will help in identifying the non-salient objects from the image for security and surveillance purposes with pin-point accuracy.

Co-saliency detection with intra-group two-stage group semantics propagation and inter-group contrastive learning

Co-salient object detection (CoSOD) aims at discovering the salient objects with repeatability from multiple relevant images and suppressing non-repetitively appearing salient and non-salient ones. Recently, learning-based methods have shown the advantages in the CoSOD tasks. However, modeling the intra-group semantics and suppressing the inter-group noise objects are still two challenges. In this paper, we propose a unified Two-stage grOup semantics PropagatIon and Contrastive learning NETwork (TopicNet) for CoSOD. TopicNet consists of an intra-group two-stage group semantics propagation module (TGSP) and an inter-group contrastive learning module (CLM). On one hand, in the first stage of TGSP, we put forward an image-to-group propagation module (IGP) to capture the inter-image similarity feature interactions. In the second stage of TGSP, a group-to-pixel propagation module (GPP) is designed to construct the connection between inter-image interactions and inter-pixel correspondences to distill consensus representation. GPP redistributes the interactions with different attention weights to each pixel so as to improve the robustness of group semantics. On the other hand, in CLM, using the design of positive samples enhances the semantic consistency of group semantics. Using the design of negative samples explores the interactive suppression relationships of co-salient objects and noise objects, which suppresses non-repetitively appearing salient and non-salient objects. TGSP and CLM are jointly integrated into an end-to-end unified network predicting the co-salient objects. Experimental results on three prevailing benchmarks reveal that TopicNet outperforms other competitors under various evaluation metrics.

Saliency Detection with Bilateral Absorbing Markov Chain Guided by Depth Information.

The effectiveness of depth information in saliency detection has been fully proved. However, it is still worth exploring how to utilize the depth information more efficiently. Erroneous depth information may cause detection failure, while non-salient objects may be closer to the camera which also leads to erroneously emphasis on non-salient regions. Moreover, most of the existing RGB-D saliency detection models have poor robustness when the salient object touches the image boundaries. To mitigate these problems, we propose a multi-stage saliency detection model with the bilateral absorbing Markov chain guided by depth information. The proposed model progressively extracts the saliency cues with three level (low-, mid-, and high-level) stages. First, we generate low-level saliency cues by explicitly combining color and depth information. Then, we design a bilateral absorbing Markov chain to calculate mid-level saliency maps. In mid-level, to suppress boundary touch problem, we present the background seed screening mechanism (BSSM) for improving the construction of the two-layer sparse graph and better selecting background-based absorbing nodes. Furthermore, the cross-modal multi-graph learning model (CMLM) is designed to fully explore the intrinsic complementary relationship between color and depth information. Finally, to obtain a more highlighted and homogeneous saliency map in high-level, we structure a depth-guided optimization module which combines cellular automata and suppression-enhancement function pair. This optimization module refines the saliency map in color space and depth space, respectively. Comprehensive experiments on three challenging benchmark datasets demonstrate the effectiveness of our proposed method both qualitatively and quantitatively.

Weight modulation in top–down computational model for target search

Computer vision research aims at building models which mimic human systems. The recent development in visual information have been used to derive computational models which address a variety of applications. Biological models help to identify the salient objects in the image. But, the identification of non-salient objects in a heterogeneous environment is a challenging task that requires a better understanding of the visual system. In this work, a weight modulation based top-down model is proposed that integrates the visual features that depend on its importance for the target search application. The model is designed to learn the optimal weights such that it biases the features of the target from the other surrounding regions. Experimental analysis is performed on various scenes on a standard dataset with the selected object in the scene. Metrics such as area under curve, average hit number and correlation reveal that the method is more suitable in target identification, by suppressing the other region.

Hybrid-Attention Network for RGB-D Salient Object Detection

Depth information has been widely used to improve RGB-D salient object detection by extracting attention maps to determine the position information of objects in an image. However, non-salient objects may be close to the depth sensor and present high pixel intensities in the depth maps. This situation in depth maps inevitably leads to erroneously emphasize non-salient areas and may have a negative impact on the saliency results. To mitigate this problem, we propose a hybrid attention neural network that fuses middle- and high-level RGB features with depth features to generate a hybrid attention map to remove background information. The proposed network extracts multilevel features from RGB images using the Res2Net architecture and then integrates high-level features from depth maps using the Inception-v4-ResNet2 architecture. The mixed high-level RGB features and depth features generate the hybrid attention map, which is then multiplied to the low-level RGB features. After decoding by several convolutions and upsampling, we obtain the final saliency prediction, achieving state-of-the-art performance on the NJUD and NLPR datasets. Moreover, the proposed network has good generalization ability compared with other methods. An ablation study demonstrates that the proposed network effectively performs saliency prediction even when non-salient objects interfere detection. In fact, after removing the branch with high-level RGB features, the RGB attention map that guides the network for saliency prediction is lost, and all the performance measures decline. The resulting prediction map from the ablation study shows the effect of non-salient objects close to the depth sensor. This effect is not present when using the complete hybrid attention network. Therefore, RGB information can correct and supplement depth information, and the corresponding hybrid attention map is more robust than using a conventional attention map constructed only with depth information.

An efficient modification of generalized gradient vector flow using directional contrast for salient object detection and intelligent scene analysis

In the field of computer vision, scene analysis is a very important area of study. To analyze the scene present in an image, in this paper, the attempt is to enhance salient object information with background information. For this, the Generalized Gradient Vector Flow model is modified by adding contrast information. The contrast information of an image is obtained by computing the Minimum Directional Contrast of the image. Using Minimum Directional Contrast as a determinant of the salient object arises from the fact that the salient objects have higher Minimum Directional Contrast than the non-salient objects. The Minimum Directional Contrast information is added to the data term of Generalized Gradient Vector Flow so that for producing contours, not only edge information is utilized, but saliency information is also used. The result gives us the salient object and added relevant background information. The algorithm is tested on three public datasets. The evaluation is done based on precision, recall, accuracy, and F1-score after comparing with six state-of-the-art methods.

The contribution of visual attention and declining verbal memory abilities to age-related route learning deficits

Our ability to learn unfamiliar routes declines in typical and atypical ageing. The reasons for this decline, however, are not well understood. Here we used eye-tracking to investigate how ageing affects people’s ability to attend to navigationally relevant information and to select unique objects as landmarks. We created short routes through a virtual environment, each comprised of four intersections with two objects each, and we systematically manipulated the saliency and uniqueness of these objects. While salient objects might be easier to memorise than non-salient objects, they cannot be used as reliable landmarks if they appear more than once along the route. As cognitive ageing affects executive functions and control of attention, we hypothesised that the process of selecting navigationally relevant objects as landmarks might be affected as well. The behavioural data showed that younger participants outperformed the older participants and the eye-movement data revealed some systematic differences between age groups. Specifically, older adults spent less time looking at the unique, and therefore navigationally relevant, landmark objects. Both young and older participants, however, effectively directed gaze towards the unique and away from the non-unique objects, even if these were more salient. These findings highlight specific age-related differences in the control of attention that could contribute to declining route learning abilities in older age. Interestingly, route-learning performance in the older age group was more variable than in the young age group with some older adults showing performance similar to the young group. These individual differences in route learning performance were strongly associated with verbal and episodic memory abilities.

Self-Association and Attentional Processing Regarding Perceptually Salient Items

Earlier work has demonstrated that attention is indirectly cognitively malleable by processes of self-association – processes by which agents explicitly associate an item with the self. We extend this work by considering the manipulation of attention to both salient and non-salient objects. We demonstrate that self-association impacts attentional processing not only of non-salient objects (i.e., shapes), but also regarding salient items known to command attention (i.e., images of food). This result indicates the flexibility and susceptibility of attentional processing to cognitive manipulation.

Attention Embedded Spatio-Temporal Network for Video Salient Object Detection

The main challenge in video salient object detection is how to model object motion and dramatic changes in appearance contrast. In this work, we propose an attention embedded spatio-temporal network (ASTN) to adaptively exploit diverse factors that influence dynamic saliency prediction within a unified framework. To compensate for object movement, we introduce a flow-guided spatial learning (FGSL) module to directly capture effective motion information in the form of attention based on optical flows. However, optical flow represents the motion information of all moving objects, including movements of non-salient objects caused by large camera motion and subtle changes in background. Therefore, using the flow-guided attention map alone causes the spatial saliency to be influenced by all moving objects rather than just the salient objects, resulting in unstable and temporally inconsistent saliency maps. To further enhance the temporal coherence, we develop an attentive bidirectional gated recurrent unit (AB-GRU) module to adaptively exploit sequential feature evolution. With this AB-GRU, we can further refine the spatiotemporal feature representation by incorporating an accommodative attention mechanism. Experimental results demonstrate that our model achieves superior empirical performance on video salient object detection. Moreover, an experiment on the extended application to unsupervised video object segmentation further demonstrates the generalization ability and stability of our proposed method.

Social comparison modulates reward-driven attentional capture.

It is well established that attention can be captured by task irrelevant and non-salient objects associated with value through reward learning. However, it is unknown whether social comparison influences reward-driven attentional capture. The present study created four social contexts to examine whether different social comparisons modulate the reward-driven capture of attention. The results showed that reward-driven attentional capture varied with different social comparison conditions. Most prominently, reward-driven attentional capture is dramatically reduced in the disadvantageous social comparison context, in which an individual is informed that the other participant is earning more monetary reward for performing the same task. These findings suggest that social comparison can affect the reward-driven capture of attention.

A color and shape based algorithm for segmentation of white blood cells in peripheral blood and bone marrow images

Computer-based imaging systems are becoming important tools for quantitative assessment of peripheral blood and bone marrow samples to help experts diagnose blood disorders such as acute leukemia. These systems generally initiate a segmentation stage where white blood cells are separated from the background and other nonsalient objects. As the success of such imaging systems mainly depends on the accuracy of this stage, studies attach great importance for developing accurate segmentation algorithms. Although previous studies give promising results for segmentation of sparsely distributed normal white blood cells, only a few of them focus on segmenting touching and overlapping cell clusters, which is usually the case when leukemic cells are present. In this article, we present a new algorithm for segmentation of both normal and leukemic cells in peripheral blood and bone marrow images. In this algorithm, we propose to model color and shape characteristics of white blood cells by defining two transformations and introduce an efficient use of these transformations in a marker-controlled watershed algorithm. Particularly, these domain specific characteristics are used to identify markers and define the marking function of the watershed algorithm as well as to eliminate false white blood cells in a postprocessing step. Working on 650 white blood cells in peripheral blood and bone marrow images, our experiments reveal that the proposed algorithm improves the segmentation performance compared with its counterparts, leading to high accuracies for both sparsely distributed normal white blood cells and dense leukemic cell clusters.

Prioritized Spatial Updating in the Intrinsic Frame of Reference

In order to maintain valid situation awareness, people need to update the spatial representations of their surroundings as objects, including themselves, move. The present study investigates the properties of spatial updating in the intrinsic frame of reference, where spatial relations are represented with respect to an external object (other than the viewer self) with an intrinsic reference direction. Three experiments were conducted using a task of direction pointing. It was found that given a relatively stable intrinsic reference direction, responses to a small number of salient objects were faster than responses to non-salient objects (Experiment 1 and Experiment 3). The salience effect disappeared when the intrinsic reference direction was no longer stable (Experiment 2). Furthermore, all three experiments revealed a type of orientation dependence similar to that found in egocentric spatial updating. Our results indicate that spatial updating in the intrinsic reference system can be easy only if a fixed reference direction is maintained and the number of objects that need to be tracked is limited.

Conformity effects in memory as a function of group size, dissenters and uncertainty

AbstractIn two experimental studies we investigated how false memory reports provided by anonymous group members affect individuals' recognition. Participants were presented with a series of salient and non‐salient objects and recognition memory was subsequently assessed for presented and non‐presented stimuli. In this recognition test, participants were confronted with (in)correct responses from bogus group members. The findings indicate that the bogus answers influenced memory reports only in the absence of a clear recollection and if the stimulus was not judged to be particularly memorable. Moreover, we found no evidence that motivational influences affected participants' judgements. However, the influence of the bogus group decreased with the presence of dissenters who broke the unanimity of the group, and it increased with higher uncertainty. Copyright © 2002 John Wiley & Sons, Ltd.