Reverse Attention for Salient Object Detection

Chen, Shuhan; Tan, Xiuli; Wang, Ben; Hu, Xuelong

doi:10.1007/978-3-030-01240-3_15

Cited by 518 publications

(365 citation statements)

References 47 publications

Supporting

Mentioning

362

Contrasting

Unclassified

Order By: Relevance

“…Multi-level feature aggregation from deep networks is also explored for detecting and refining the detection [17,44,13]. Recent works apply attention mechanisms for learning global and local contexts [21] or learning foreground/background attention maps [8] to help detect salient objects and eliminate non-salient objects.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Where Is My Mirror?

Yang

Mei

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

Mirrors are everywhere in our daily lives. Existing computer vision systems do not consider mirrors, and hence may get confused by the reflected content inside a mirror, resulting in a severe performance degradation. However, separating the real content outside a mirror from the reflected content inside it is non-trivial. The key challenge is that mirrors typically reflect contents similar to their surroundings, making it very difficult to differentiate the two. In this paper, we present a novel method to segment mirrors from an input image. To the best of our knowledge, this is the first work to address the mirror segmentation problem with a computational approach. We make the following contributions. First, we construct a large-scale mirror dataset that contains mirror images with corresponding manually annotated masks. This dataset covers a variety of daily life scenes, and will be made publicly available for future research. Second, we propose a novel network, called Mirror-Net, for mirror segmentation, by modeling both semantical and low-level color/texture discontinuities between the contents inside and outside of the mirrors. Third, we conduct extensive experiments to evaluate the proposed method, and show that it outperforms the carefully chosen baselines from the state-of-the-art detection and segmentation methods.

show abstract

Section: Related Workmentioning

confidence: 99%

“…well by the state-of-the-art segmentation methods [47,12]. Meanwhile, as the contents reflected by the mirrors may not necessarily be salient, directly applying state-of-the-art saliency detection methods [8,21] for detecting mirrors is also not appropriate.…”

Section: Introductionmentioning

confidence: 99%

Where Is My Mirror?

Yang

Mei

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

show abstract

“…We compare the proposed saliency detection method against previous 18 state-of-the-art methods, namely, MDF [13], RFCN [31], DHS [32], UCF [46], Amulet [34], NLDF [47], DSS [48], RAS [49], BMPM [33], PAGR [50], PiCANet [51], SRM [18], DGRL [17], MLMS [52], AFNet [53], CapSal [54], BASNet [55], and CPD [16]. We perform comparisons on five challenging datasets.…”

Section: Comparison With the State-of-the-artmentioning

confidence: 99%

DFNet: Discriminative feature extraction and integration network for salient object detection

Noori

Mohammadi

Majelan

et al. 2020

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

Despite the powerful feature extraction capability of Convolutional Neural Networks, there are still some challenges in saliency detection. In this paper, we focus on two aspects of challenges: i) Since salient objects appear in various sizes, using single-scale convolution would not capture the right size. Moreover, using multi-scale convolutions without considering their importance may confuse the model. ii) Employing multilevel features helps the model use both local and global context. However, treating all features equally results in information redundancy. Therefore, there needs to be a mechanism to intelligently select which features in different levels are useful. To address the first challenge, we propose a Multi-scale Attention Guided Module. This module not only extracts multi-scale features effectively but also gives more attention to more discriminative feature maps corresponding to the scale of the salient object. To address the second challenge, we propose an Attention-based Multi-level Integrator Module to give the model the ability to assign different weights to multi-level feature maps. Furthermore, our Sharpening Loss function guides our network to output saliency maps with higher certainty and less blurry salient objects, and it has far better performance than the Cross-entropy loss. For the first time, we adopt four different backbones to show the generalization of our method. Experiments on five challenging datasets prove that our method achieves the state-of-the-art performance. Our approach is fast as well and can run at a real-time speed.

show abstract

“…Zhang et al [53] introduce a multi-path recurrent feedback scheme to progressively enhance the saliency prediction map. RA [4] introduces reverse attention with side-output residual learning to refine the saliency map in a top-down manner. Also, skip connections are widely applied to integrate prediction maps from CNNs [51,11].…”

Section: Related Workmentioning

confidence: 99%

“…2) By adding message-passing between Table 3. Evaluation results on six dataset and with models DRFI [14], MDF [22], RFCN [40], DHS [26], Amulet [51], UCF [52], DCL [23], MSR [21], DSS [11], RA [4] and the deep unified CRF model. "+" marks the models utilizing Dense-CRF [19] for post-processing.…”

Section: Model Analysismentioning

confidence: 99%

Structured Modeling of Joint Deep Feature and Prediction Refinement for Salient Object Detection

Hong

et al. 2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

View full text Add to dashboard Cite

Recent saliency models extensively explore to incorporate multi-scale contextual information from Convolutional Neural Networks (CNNs). Besides direct fusion strategies, many approaches introduce message-passing to enhance CNN features or predictions. However, the messages are mainly transmitted in two ways, by feature-tofeature passing, and by prediction-to-prediction passing. In this paper, we add message-passing between features and predictions and propose a deep unified CRF saliency model . We design a novel cascade CRFs architecture with CNN to jointly refine deep features and predictions at each scale and progressively compute a final refined saliency map. We formulate the CRF graphical model that involves message-passing of feature-feature, feature-prediction, and prediction-prediction, from the coarse scale to the finer scale, to update the features and the corresponding predictions. Also, we formulate the mean-field updates for joint end-to-end model training with CNN through back propagation. The proposed deep unified CRF saliency model is evaluated over six datasets and shows highly competitive performance among the state of the arts.

show abstract

Reverse Attention for Salient Object Detection

Cited by 518 publications

References 47 publications

Where Is My Mirror?

Where Is My Mirror?

DFNet: Discriminative feature extraction and integration network for salient object detection

Structured Modeling of Joint Deep Feature and Prediction Refinement for Salient Object Detection

Contact Info

Product

Resources

About