Salient object detection network with multi-scale feature refinement and boundary feedback

Many existing salient object detection methods are dedicated to fusing features from different levels of a pre-trained convolutional neural network (CNN). However, these methods can easily lead to internal discontinuities within the salient objects because of unreasonable feature fusion strategies and short-range dependencies resulting from common convolution and pooling operations. In this paper, we propose a novel non-local duplicate pooling (NLDP) network to overcome these internal discontinuities. NLDP begins by removing the first few convolutional layers of a classic CNN, which have small receptive fields and require large amounts of calculation. A novel duplicate pooling module (DPM) is then used to generate richer and more detailed saliency maps. This is achieved by constructing a double-pathway that can integrating partial feature maps. Within the DPM, a non-local module (NLM) is used to obtain long-range dependencies. This enhances the internal continuities between the saliency maps. Comprehensive experiments conducted on six benchmark datasets have confirmed the increased effectiveness and detection speed of our method in relation to other salient object detection methods.

A pooling-based feature pyramid network for salient object detection

Salient object detection (SOD) has achieved remarkable performance in well‐illuminated scenarios. However, in inevitable low‐light environments, SOD faces inherent challenges such as insufficient illumination, noise interference, and contrast degradation, which severely compromise the perceptual accuracy of visual models. To address these issues, we propose a novel Edge‐Guided Dual‐Stream Collaborative Network (EGDSNet), which enhances the integrity of salient objects through multi‐level feature interaction between edge detection and segmentation branches. Specifically, EGDSNet adopts a dual‐branch architecture: the edge detection branch integrates a Local Context Refinement (LCR) module and a Cross‐Modal Edge Refinement (CER) module to suppress noise while reinforcing boundary responses of targets. The segmentation branch employs a Global Feature Fusion (GFF) module and an Edge‐Guided Feature Enhancement (EFE) module to achieve complementary integration of global semantics and local details. The two branches interact via an edge‐guided attention mechanism, enabling saliency predictions to maintain structural integrity while attaining precise boundary localization. Additionally, the Residual Channel Attention (RCA) module is leveraged to further suppress noise and extract structural details from low‐level features. Comprehensive experiments on the LLI and SOD‐LL datasets demonstrate that our model outperforms state‐of‐the‐art methods in both quantitative metrics and qualitative visualizations, validating its superiority in low‐light SOD tasks.

Given that existing salient object detection methods cannot effectively predict the fine contours of salient objects when extracting local or global contexts and features, we propose a novel contour self-compensated network (CSCNet) to generate a more accurate saliency map with complete contour. Unlike the common binary saliency detection, we reconstruct the salient object detection problem into a multi-classification problem of the background, the salient object, and the salient contour, where the salient contour is used as the third label for ground truth. Meanwhile, the image and its superpixel map are concatenated as the input of our network to add more edge information. Also, a penalty loss is proposed to restrict the spatial relationship between the background, objects, and their contours. Experimentally, we evaluate the proposed CSCNet on six benchmark datasets in both accuracy and efficiency and evaluate the attribute-based performance on the SOC dataset. Compared with 13 state-of-the-art algorithms, our CSCNet can detect salient objects more accurately and completely without adding too many convolutional layers and parameters.

Attention and boundary guided salient object detection

Salient object detection (SOD) aims to determine the most visually attractive objects in an image. With the development of virtual reality (VR) technology, 360° omnidirectional image has been widely used, but the SOD task in 360° omnidirectional image is seldom studied due to its severe distortions and complex scenes. In this article, we propose a multi-projection fusion and refinement network (MPFR-Net) to detect the salient objects in 360° omnidirectional image. Different from the existing methods, the equirectangular projection (EP) image and four corresponding cube-unfolding (CU) images are embedded into the network simultaneously as inputs, where the CU images not only provide supplementary information for EP image but also ensure the object integrity of cube-map projection. In order to make full use of these two projection modes, a dynamic weighting fusion (DWF) module is designed to adaptively integrate the features of different projections in a complementary and dynamic manner from the perspective of inter and intrafeatures. Furthermore, in order to fully explore the way of interaction between encoder and decoder features, a filtration and refinement (FR) module is designed to suppress the redundant information of the feature itself and between the features. Experimental results on two omnidirectional datasets demonstrate that the proposed approach outperforms the state-of-the-art methods both qualitatively and quantitatively. The code and results can be found from the link of https://rmcong.github.io/proj_MPFRNet.html.

Recent progress in salient object detection (SOD) mainly depends on the Atrous Spatial Pyramid Pooling (ASPP) module for multi-scale learning. Intuitively, different input images, different pixels, and different network layers may have different preferences for various feature scales. However, ASPP treats all feature scales as equally important by a simple sum operation. To this end, we propose Attentive Atrous Spatial Pyramid Pooling (A2SPP) by adding a new Cubic Information-Embedding Attention (CIEA) module at each branch of ASPP. In this way, each position in the 3D feature map can automatically learn the feature scales it prefers. Specifically, CIEA consists of Spatial-Embedding Channel Attention (SECA) and Channel-Embedding Spatial Attention (CESA). Instead of the previous direct squeeze and ignoring of one dimension when computing the attention for the other dimension, SECA/CESA attempts to embed spatial/channel information into channel/spatial attention, respectively. In addition, CIEA learns SECA and CESA for each 3D position simultaneously rather than previous separate computation of channel and spatial attention for each 2D position. Incorporating A2SPP and CIEA, the proposed A2SPPNet performs favorably against previous state-of-the-art SOD methods.

Effectively capturing the multi-scale, multi-channel, and boundary features of target images is beneficial for improving the accuracy of salient object detection. A multichannel feature enhancement module (MFEM) is constructed by concatenating two feature maps along the channel, performing convolution activation, dot product addition, and convolution operations. An edge feature enhancement module (EFEM) is constructed by performing an inverse attention mechanism and Gaussian-Laplace transform on the predicted map, channel-wise concatenation, and convolution operations with the feature submap. Multi-scale features are extracted using a SwinTransformer encoder and enhanced using the MFEM. The enhanced features are then fused using a cross-attention mechanism for multi-level features. The fused feature map, the high-frequency feature map obtained by Gaussian-Laplace transforming the predicted map, and the external edge features are input into the EFEM to obtain the salient object detection results. This results in a novel salient object detection model with multi-scale, multi-channel, and boundary feature enhancement: MfENet. Experimental results show that the Mean Absolute Error (MAE), S-measure and F-measure accuracies of our method on the HKU-IS dataset are 0.018,0.924 and 0.959 respectively. These results demonstrate the effectiveness of our method. Furthermore, this work highlights the importance of multi-scale and boundary awareness in enhancing salient object detection, and its contribution to the technical field relevant to the theme of this conference.

While salient object detection (SOD) on 2D images has been extensively studied, there is very little SOD work on 3D measurement surfaces. We propose an effective point transformer-based SOD network for 3D measurement point clouds, termed PSOD-Net. PSOD-Net is an encoder-decoder network that takes full advantage of transformers to model the contextual information in both multi-scale point- and scene-wise manners. In the encoder, we develop a Point Context Transformer (PCT) module to capture region contextual features at the point level; PCT contains two different transformers to excavate the relationship among points. In the decoder, we develop a Scene Context Transformer (SCT) module to learn context representations at the scene level; SCT contains both Upsampling-and-Transformer blocks and Multi-context Aggregation units to integrate the global semantic and multi-level features from the encoder into the global scene context. Experiments show clear improvements of PSOD-Net over its competitors and validate that PSOD-Net is more robust to challenging cases such as small objects, multiple objects, and objects with complex structures. Code is available at: https://github.com/ZeyongWei/PSOD-Net.

Beauty and Personal care product retrieval has attracted more and more attention due to its wide application value. However, due to the diversity of data and the complexity of image background, this task is very challenging. In this paper, we propose a multi-feature fusion method based on salient object detection to improve retrieval performance. The key of our method is to extract the foreground objects of the query set by using the salient object detection network, so as to eliminate the background interference. Then the foreground target images and dataset are put into the multi-classification networks to extract multiple fusion features for retrieval. We use the perfect-500k dataset for experiments, and the results show that our method is effective. Our method ranked 2st in the Grand Challenge of AI Meets Beauty in ACM Multimedia 2020 with a MAP score of 0.43729. We released our code on GitHub:github.com/R-M-Yan/ACMMM2020AIMeetBeauty.

The performance of the salient object detection of strip surface defects has been promoted largely by deep learning based models. However, due to the complexity of strip surface defects, the existing models perform poorly in the challenging scenes such as noise disturbance, and low contrast between defect regions and background. Meanwhile, the detection results of existing models often suffer from coarse boundary details. Therefore, we propose a novel saliency model, namely an Edge-aware Multi-level Interactive Network, to detect the defects from the strip steel surface. Concretely, our model adopts the U-shape architecture where the two crucial points are the interactive feature integration and the edge-guided saliency fusion. Firstly, except the skip connection that combines the same stage of encoder and decoder, we deploy another connection, where the features from adjacent levels of encoder are transferred to the same stage of decoder. By this way, we are able to provide an effective fusion of multi-level deep features, yielding a well depiction for defects. Secondly, to give well-defined boundaries for prediction results, we add the edge extraction branch after each decoder block, where the progressive feature aggregation endows the edge with precise details and complete object cues. Meanwhile, together with the edge extraction branches, we deploy the saliency prediction branch at each decoder stage. After that, coupled with the fine edge information, we fuse all outputs of saliency prediction branches into the final saliency map, where the edge cue steers the saliency result to pay more attention to the boundary details. Following this way, we can provide a high-quality saliency map which can accurately locate and segment the defects. Extensive experiments are performed on the public dataset, and the results prove the effectiveness and robustness of our model which consistently outperforms the state-of-the-art models.

Point cloud salient object detection has attracted the attention of researchers in recent years. Since existing works do not fully utilize the geometry context of 3D objects, blurry boundaries are generated when segmenting objects with complex backgrounds. In this paper, we propose a geometry-aware 3D salient object detection network that explicitly clusters points into superpoints to enhance the geometric boundaries of objects, thereby segmenting complete objects with clear boundaries. Specifically, we first propose a simple yet effective superpoint partition module to cluster points into superpoints. In order to improve the quality of superpoints, we present a point cloud class-agnostic loss to learn discriminative point features for clustering superpoints from the object. After obtaining superpoints, we then propose a geometry enhancement module that utilizes superpoint-point attention to aggregate geometric information into point features for predicting the salient map of the object with clear boundaries. Extensive experiments show that our method achieves new state-of-the-art performance on the PCSOD dataset.

Global and local information aggregation network for edge-aware salient object detection

There are two main areas in saliency detection: fixation prediction and salient object detection. Researchers always use bottom-up low-level features and straightforward rules for salient object detection. However these rules are unreliable in common scenes. As a result, most salient object detection technologies can gain a fine performance in limited situation. While in a more complex scene (such as PASCAL-S dataset), they may miss the really wanted salient objects. Fixation prediction is a technology of predicting the regions which can attract human's attention. Most fixation prediction technologies are based on the discoveries of neuroscience and cognitive psychology. Their rules for fixation prediction more accord with human's sensory mechanism. Thus, fixation map can provide a more reliable salient priori. Based on above reasons, we propose a novel salient object detection method with fixation priori. We refine fixation map to gain a reliable priori, and conduct salient object detection based on the priori. We perform experiment on two datasets, and find that our method outperforms the comparative state-of-art salient object detection technologies.

Due to the rapid development of deep learning, salient object detection has achieved promising results in recent years. However, how to construct a powerful saliency detection network to generate saliency maps with completely highlighted salient objects and effectively suppressed background noise still remains an open and challenging problem. In this paper, we propose a novel trifurcated cascaded refinement network (TCRNet) to explore multi-level feature fusion and global information representation. Specifically, all the side-output features from the backbone network are divided into three branches (i.e., global information, deep information and shallow information) according to feature similarities, and then the global contexts, semantic knowledge and fined details respectively generated by the three branches are integrated in a cascaded refinement way. This refinement strategy aims to guide the feature learning of important regions and suppress background noises. Moreover, we propose a cross-level feature fusion (CFF) module to fuse multi-level similar features within a same branch by exploring their interrelationship, thus capturing the important semantic information and structure details. To this end, the complementarity of similar features is utilized to refine feature at each level by sharing an encoded spatial weight map. Besides, we design a context-aware feature learning (CFL) module to capture global contexts in a single layer by taking account of both single-feature representation and multi-feature interaction, thus providing important location information of salient objects. Extensive experiments on five benchmarks are conducted to demonstrate that the proposed network TCRNet performs favorably against 20 state-of-the-art salient object detection methods on five benchmark datasets, indicating the effectiveness and superiority of our method.