A Simple Pooling-Based Design for Real-Time Salient Object Detection

By Jiangjiang Liu, Qibin Hou, Ming-Ming Cheng, Jiashi Feng, Jianmin Jiang. 2019

991 citations recorded by the source corpus.

We solve the problem of salient object detection by investigating how to expand the role of pooling in convolutional neural networks. Based on the U-shape architecture, we first build a global guidance module (GGM) upon the bottom-up pathway, aiming at providing layers at different feature levels the location information of potential salient objects. We further design a feature aggregation module (FAM) to make the coarse-level semantic information well fused with the fine-level features from the top-down path- way. By adding FAMs after the fusion operations in the top-down pathway, coarse-level features from the GGM can be seamlessly merged with features at various scales. These two pooling-based modules allow the high-level semantic features to be progressively refined, yielding detail enriched saliency maps. Experiment results show that our proposed approach can more accurately locate the salient objects with sharpened details and hence substantially improve the performance compared to the previous state-of-the-arts. Our approach is fast as well and can run at a speed of more than 30 FPS when processing a 300×400 image. Code can be found at http://mmcheng.net/poolnet/.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning (2021, neutral): “connections [11] and pooling-based modules [21].”
Strip Pooling: Rethinking Spatial Pooling for Scene Parsing (2020, methodology): “Inspired by [34, 35], a convolutional layer is added after the fusion operation in each sub-module to reduce the aliasing effect brought by downsampling operations.”
Res2Net: A New Multi-Scale Backbone Architecture (2019, background): “Multi-scale feature representations of CNNs are of great importance to a number of vision tasks including object detection [53], face analysis [4], [51], edge detection [45], semantic segmentation [6], salient object detection [42], [78], and skeleton detection [80], boosting the model performance o”
BBS-Net: RGB-D Salient Object Detection with a Bifurcated Backbone Strategy Network (2020, methodology): “re typically based on handcrafted features and fall short in capturing high-level semantic information (see also [14], [15]). Recently, convolutional neural networks (CNNs) have been used for RGB SOD [16], [17], achieving better performance compared to the traditional methods. However, the performan”
A Multi-Task Mean Teacher for Semi-Supervised Shadow Detection (2020, background): “Recent saliency detectors [14, 15] have also proved this point, in which edge knowledge is helpful to improve the saliency detection quality.”
Zoom In and Out: A Mixed-scale Triplet Network for Camouflaged Object Detection (2022, methodology): “Thus, some recent CNN-based COD methods [9,31,43,59] and SOD methods [15,23,34,35,66,67,69] explore the combination strategy of inter-layer features to enhance the feature representation.”
Uncertainty-Guided Transformer Reasoning for Camouflaged Object Detection (2021, methodology): “Figure 6: Qualitative comparisons between different models: (c) UGTR, (d) SINet [12], (e) EGNet [59], (f) POOLNet [33], and (g) PFANet [60].”
Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation (2020, methodology): “entation network. For pseudo groundtruth generation, we follow current popular pipeline [20,22,23, 31,32,81], that uses localization maps to extract class-specic object cues and adopts saliency maps [19,38] to get background cues. For the semantic segmentation network, as in [20,23,31,32], we choose”
Context-aware Cross-level Fusion Network for Camouflaged Object Detection (2021, background): “PoolNet [Liu et al., 2019], which is equipped with a pyramid pool module, employs a global guidance module to obtain the location information of underlying salient targets for different feature layers.”
Label Decoupling Framework for Salient Object Detection (2020, methodology): “To demonstrate the effectiveness of the proposed method, 14 state-of-the-art SOD methods are introduced to compare, including BMPM [38], DGRL [28], R(3)Net [7], RAS [4], PiCA-R [21], AFNet [13], BASNet [23], CPD-R [33], EGNet-R [41], PAGE [31], TDBU [30], SCRN [34], SIBA [24] and PoolNet [20].”
Uncertainty-aware Joint Salient Object and Camouflaged Object Detection (2021, background): “Salient Object Detection Models Existing deep saliency detection models [47, 48, 42, 29, 46, 12, 38] are mainly designed to achieve structure-preserving saliency predictions.”
F3Net: Fusion, Feedback and Focus for Salient Object Detection (2019, methodology): “(Liu et al. 2019) utilized simple pooling and feature aggregation module to build fast and high performance model.”