Learning to Detect Salient Objects with Image-Level Supervision

By Lijun Wang, Huchuan Lu, Mengyang Feng, Dong Wang, Baocai Yin, Xiang Ruan. 2017

1,257 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, 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 Cognitive Neuroscience

In Computer Vision and Pattern Recognition

Citations in the corpus

FocalClick: Towards Practical Interactive Image Segmentation (2022, methodology): “We also report the performance of our model trained on SBD [15], and a large combined dataset [2, 7, 13, 24, 26, 39, 45, 48].”
A Bi-Directional Message Passing Model for Salient Object Detection (2018, methodology): “DUTS [28] is a large scale dataset, which contains 10553 images for training and 5019 images for testing.”
Visual Saliency Transformer (2021, neutral): “Settings DUTS [25] HKU-IS [10] PASCAL-S [12] SOD [”
Simultaneously Localize, Segment and Rank the Camouflaged Objects (2021, background): “[63] introduced the class activation map (CAM) to estimate the discriminative region of each class, which is the basis of many weakly supervised methods [1, 51, 17, 50, 24, 43, 47].”
Amulet: Aggregating Multi-level Convolutional Features for Salient Object Detection (2017, background): “…features at each resolution and predicts saliency maps in a recursive manner. dataset [5]) achieves new state-of-the-art performance on other large-scale salient object detection datasets, including the recent DUTS [42], DUT-OMRON [47], ECSSD [46], HKU-IS [50], PASCAL-S [26], SED [2] and SOD [46].”
Interactive Two-Stream Decoder for Accurate and Fast Saliency Detection (2020, methodology): “F-score and FPS comparisons of our ITSD with AFNet [8], PCA [22], Amulet [45], EGNet [48], CKT [18], CTLoss [4], NLDF [24], PoolNet [21], PAGE [35] and CPD [37] on DUTS-TE dataset [31].”
Stacked Cross Refinement Network for Edge-Aware Salient Object Detection (2019, methodology): “We evaluate the proposed approach on six traditional benchmark datasets: ECSSD [33], PASCAL-S [21], DUTS [28], HKU-IS [19], DUT-OMRON [34] and THUR 15 K [6].”
AnyDoor: Zero-shot Object-level Image Customization (2023, neutral): “MSRA-10K [4] Single Image 10,000 ✗ High DUT [50] Single Image 15,572 ✗ High HFlickr [15] Single Image 4,833 ✗ High LVIS [20] Single Image 118,287 ✗ High SAM (subset) [27] Single Image 100,864 ✗ High”
Rethinking Space-Time Networks with Improved Memory Coverage for Efficient Video Object Segmentation (2021, methodology): “Following previous practices [18, 21], we ﬁrst pretrain the model on static image datasets [73, 74, 75, 76, 77] with synthetic deformation then perform main training on YouTubeVOS [78] and DAVIS [3, 65].”
Cascaded Partial Decoder for Fast and Accurate Salient Object Detection (2019, methodology): “Method Backbone FPS ECSSD [39] HKU-IS [16] DUT-OMRON [40] DUTS [33] PASCAL-S [19] maxF avgF MAE maxF avgF MAE maxF avgF MAE maxF avgF MAE maxF avgF MAE”
A Simple Pooling-Based Design for Real-Time Salient Object Detection (2019, methodology): “(a) PASCAL-S [21] (b) HKU-IS [18] (c) DUTS-TE [35]”
Saliency Detection via a Multiple Self-Weighted Graph-Based Manifold Ranking (2020, methodology): “We employ DUTS-test [54] which contains 5,019 images collected from the ImageNet test set and the SUN data set.”
Anabranch network for camouflaged object segmentation (2019, methodology): “DHS and WSS reduce accuracy while DSS and SRM slightly increase accuracy (see Table 6).”
Weakly Supervised Object Localization and Detection: A Survey (2021, neutral): “The whole weakly supervised learning framework may be designed in a compact network model, such as in [46], [76], [101], [106], [117], [180], [182], [204], [219], [221], or contain several function-distinct DNN components, such as in [27], [98], [140], [160], [172], [193], [209].”
How Much Position Information Do Convolutional Neural Networks Encode? (2020, methodology): “Datasets: We use the DUT-S dataset (Wang et al., 2017) as our training set, which contains 10 , 533 images for training.”
Counterfactual Generative Networks (2021, methodology): “The U2-Net was trained for salient object detection on DUTS-TR (10553 images) (Wang et al., 2017).”
Detect Globally, Refine Locally: A Novel Approach to Saliency Detection (2018, methodology): “We evaluate the proposed framework on five popular datasets: ECSSD [31], DUTOMRON [32], THUR15K [5], HKU-IS [17], and DUTS [27].”