The DOI matches its Crossref record; 56 of the 57 references it cites were found in a registry or an index.
Hide-and-Seek: Forcing a Network to be Meticulous for Weakly-Supervised Object and Action Localization
Open accessAbstract
We propose `Hide-and-Seek', a weakly-supervised framework that aims to improve object localization in images and action localization in videos. Most existing weakly-supervised methods localize only the most discriminative parts of an object rather than all relevant parts, which leads to suboptimal performance. Our key idea is to hide patches in a training image randomly, forcing the network to seek other relevant parts when the most discriminative part is hidden. Our approach only needs to modify the input image and can work with any network designed for object localization. During testing, we do not need to hide any patches. Our Hide-and-Seek approach obtains superior performance compared to previous methods for weakly-supervised object localization on the ILSVRC dataset. We also demonstrate that our framework can be easily extended to weakly-supervised action localization.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“However, various types of industrial image defects do not meet all the above assumptions [42,43].”
From An Improved VGG19 Transfer Learning Strip Steel Surface Defect Recognition Deep Neural Network Based on Few Samples and Imbalanced Datasets (2021) · cites this paper as [43]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
Fully convolutional networks for semantic segmentation
Long, J., Shelhamer, E., & Darrell, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298965 · Cited by 38,834
(2014) · Cited by 33,419
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
(2016) · Cited by 31,590
Girshick, R. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.169 · Cited by 28,105
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818
Visualizing and understanding convolutional networks
Zeiler, M. D., & Fergus, R. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10590-1_53 · Cited by 15,839
Flagged first, then by citations
1 of the 57 references couldn’t be matched to a known paper.
Research works printed but not linked: 5
- jiang, 2014, THUMOS Challenge Action Recognition with A Large Number of Classes
- jhuang, 2013, Towards undestanding action recognition, ICCV
- 10.1109/CVPR.2016.311 · DOI · Paper page
- cinbis, 2015, Weakly supervised object localization with multi-fold multiple instance learning
- wan, 2013, Regularization of neural network using dropconnect, ICML
Cited by
Sorted by citations
Items 601–700 of 720
Guo, Y., Liang, X., Hui, T., Wu, B., & Zheng, X. (2022) · International Conference on Multimedia Retrieval · doi:10.1145/3512527.3531360 · Cited by 1
“For example, hiding or erasing the region of an object to seek more diverse parts, however, such methods [21, 28] either hide fixed-size patches randomly or require repetitive model training and response aggregation steps.”
BackgroundWang, J., & Wang, C. (2022) · Journal of Electronic Imaging · doi:10.1117/1.jei.31.6.063019 · Cited by 1
“In response to this problem, early works adopted elaborate erasure strategies 18 – 20 or multibranch architectures; 15 both force the model to focus on different regions of the video, further discovering the entire action and achieving complete localization.”
MethodA Simple Approach and Benchmark for 21,000-Category Object Detection
Lin, Y., Li, C., Cao, Y., Zhang, Z., Wang, J., Wang, L., Liu, Z., & Hu, H. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20083-0_1 · Cited by 1
“The WSOL meth-ods [53,35,44,3,48,49,50,29,10] are mostly based on CAM [53] with the class activation maps as an indicator of the object area.”
BackgroundMethodWeakly Supervised Object Localization with Noisy-Label Learning
Fan, Y., Wei, S., Tan, C., Chen, X., & Zhao, Y. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-18916-6_39 · Cited by 1
EGSA: Enhanced and Global Semantic Activation for Weakly Supervised Object Localization
Liu, Y., Wang, L., Xu, X., & Luo, X. (2022) · IEEE International Conference on Networking, Sensing and Control (ICNSC) · doi:10.1109/icnsc55942.2022.10004147 · Cited by 1
“The HAS model [10] was proposed to divide the input image into fixed-size grids for random erasing.”
MethodShalobanov, S. V., & Shalobanov, S. S. (2023) · International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM) · doi:10.1109/icieam57311.2023.10139270 · Cited by 1
“The graphs of the obtained dependencies for different input signals for the same block have a similar shape [22-24], which allows us to conclude that their form is primarily determined by the place of the controlled block in the system [25-27].”
BackgroundA Weakly Supervised Semantic Segmentation Method on Lung Adenocarcinoma Histopathology Images
Lan, X., Mei, J., Lin, R., Chen, J., & Zhang, Y. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-4742-3_57 · Cited by 1
Context-Based Masking for Spontaneous Venous Pulsations Detection
Sheng, H., Yu, X., Li, X., & Golzan, M. (2023) · Lecture notes in computer science · doi:10.1007/978-981-99-8388-9_42 · Cited by 1
Weak-shot Semantic Segmentation Based on Encoder-Decoder Structure
Zhang, Y., Wang, Y., Wei, Y., Feng, B., Hu, X., Guo, D., & Zou, B. (2023) · International Conference on Software Engineering and Computer Science (CSECS) · doi:10.1109/csecs60003.2023.10428525 · Cited by 1
“Weakly Supervised Semantic Segmentation (WSSS) is a weakly supervised learning approach for the task of semantic segmentation [9].”
BackgroundOracle Bone Character Detection Based on Optimized YOLOv10
Wu, L., Zhen, Q., & Liu, G. (2024) · International Conference on Machine Learning, Pattern Recognition and Automation Engineering (MLPRAE) · doi:10.1145/3696687.3696701 · Cited by 1
“In the clipping methods, Hide and Seek [ 19 ] , Cutout [ 20 ] , Random Erasing [ 21 ] , and Gridmask [ 22 ] , we choose the best performing Gridmask.”
Zhang, H., Han, X., & Sun, W. (2024) · Remote Sensing · doi:10.3390/rs16244806 · Cited by 1
“…dataset and new augmented dataset can be expressed as Combined with Equations (2) and (3), we have On the other hand, there are a series of data augmentation methods that generate new augmented images by occlusion or re-editing of the original images, such as Cutout [18], RE [19], HaS [20], etc.”
Global Discriminative Information Search and Focus for SAR Target Recognition
Zhao, C., Wang, D., Zhang, S., & Kuang, G. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3552578 · Cited by 1
“Although the CAM-based methods [41], [42], [43] can localize the target area with a small number of parameters, it suffers from the partial activation problem, i.e., the activated target region is incomplete.”
Weakly supervised temporal action localization via a multimodal feature map diffusion process
Zou, Y., Zhao, Q., & Li, S.-S. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.111044 · Cited by 1
Counterfactual Co-Occurring Learning for Bias Mitigation in Weakly-Supervised Object Localization
Shao, F., Luo, Y., Chen, L., Liu, P., Yang, W., Yang, Y., & Xiao, J. (2026) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2026.3651135 · Cited by 2
Human-Like Delicate Region Erasing Strategy for Weakly Supervised Detection
En, Q., Duan, L., Zhang, Z., Bai, X., & Zhang, Y. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33013502 · Cited by 1
“Ourwork has similarities with (Wei et al. 2017; Kumar Singh and Jae Lee 2017), but we do not require the clas-sification network to be retrained after each image is erased.”
BackgroundCategory-Level Multi-Attention based Boundary Refinement for Action Detection
Dong, P., Zhu, L., & Zhang, Y. (2019) · IEEE International Conference on Signal and Image Processing · doi:10.1109/siprocess.2019.8868533 · Cited by 2
“,2017) [15] 36.”
Selecting Discriminative Features for Fine-Grained Visual Classification
Xu, Q., Li, L., Chen, Q., & Luo, B. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-41404-7_65 · Cited by 1
Discriminative Regions Erasing Strategy for Weakly-Supervised Temporal Action Localization
Huanbin, Z., Zhu, S., & Yu, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-60639-8_53 · Cited by 1
Evaluation of Inpainting and Augmentation for Censored Image Queries
Black, S., Keshavarz, S., & Souvenir, R. (2021) · International Journal of Computer Vision · doi:10.1007/s11263-020-01403-1 · Cited by 1
“There is much recent interest in the computer vision community for this approach for data augmentation in both supervised (DeVries and Taylor 2017; Singh and Lee 2017; Zhong et al. 2020) and self-supervised learning (Pathak et al.”
BackgroundMethodInfluentialContext Driven Network with Bayes for Weakly Supervised Temporal Action Localization
Yu, J., Ge, Y., Li, Z., Chen, Z., & Qin, X. (2021) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme51207.2021.9428387 · Cited by 1
“Hide-and-Seek[5] random erases video sequences during training.”
MethodUnsupervised Object Segmentation Based on Bi-Partitioning Image Model Integrated with Classification
Choi, H.-T., & Hong, B. (2021) · Electronics · doi:10.3390/electronics10182296 · Cited by 1
“[30] use randomly generated hidden image patches for each image at every iteration.”
MethodWeakly-Supervised Temporal Action Localization by Background Suppression
Liu, M., Gao, X., Ge, F., Liu, H., & Li, W. J. (2021) · Chinese Control Conference (CCC) · doi:10.23919/ccc52363.2021.9550506 · Cited by 1
“Hide-and-seek [9] allows the classification network to distribute attention to each area more evenly by randomly hiding some patches, instead of just focusing on the most discriminative regions.”
BackgroundZhu, M., Wan, S., Jin, P., & Xu, J. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87355-4_35 · Cited by 1
One-Shot Example Videos Localization Network for Weakly-Supervised Temporal Action Localization
Liu, Y., Zhang, W., Li, G., Su, L., & Huang, Q. (2021) · IEEE International Conference on Multimedia Information Processing and Retrieval (MIPR) · doi:10.1109/mipr51284.2021.00026 · Cited by 1
Adversarial Decoupling for Weakly Supervised Semantic Segmentation
Sun, G., Meng, Y., & Luo, W. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88013-2_16 · Cited by 1
Unsupervised Domain Adaptation via Attention Augmented Mutual Networks for Person Re-identification
Tian, H., & Hu, J. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-93046-2_41 · Cited by 1
Wang, L., Duali, J., Chen, Z., Chen, G., & Liu, G. (2021) · International Conference on Information, Cybernetics, and Computational Social Systems (ICCSS) · doi:10.1109/iccss53909.2021.9722008 · Cited by 1
RSMNet: A Regional Similar Module Network for Weakly Supervised Object Localization
Ling, Z., Li, L., & Zhang, A. (2022) · Neural Processing Letters · doi:10.1007/s11063-022-10849-y · Cited by 1
Bloshchinskiy, V. D., Shalobanov, S. V., & Shalobanov, S. S. (2022) · International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM) · doi:10.1109/icieam54945.2022.9787235 · Cited by 1
Explored seeds generation for weakly supervised semantic segmentation
Chow, T., Deng, H., Yang, Y., Lin, Z., Zhuang, H., & Du, S. (2023) · Neural Computing and Applications · doi:10.1007/s00521-023-09073-z · Cited by 1
“The CAM is often at the core of recent works [4–6 , 9– 16 ] on weakly supervised semantic In contrast, explicit seed learning methods consider seeds as a single unit utilized in a model loss function.”
Truong, T.-D., Duong, C. N., Quach, K. G., Nguyen, D., Le, N., Luu, K., & Bui, T. D. (2020) · arXiv (Cornell University) · Cited by 1
Explainability-aided Domain Generalization for Image ClassificationPreprint
Schmidt, R. M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.01742 · Cited by 3
“These include dropping random patches of input images (Cutout & HaS) [40, 172] or channels of the feature map (SpatialDropout) [185], dropping contiguous regions of the feature maps (DropBlock) [64], dropping features of high activations across feature maps and channels (MaxDrop) [139], or generalizing the traditional…”
BackgroundFrom medical imaging to explainable artificial intelligence
Tjoa, E. (2022) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/164493 · Cited by 1
Ravanbakhsh, E., Niu, C., Liang, Y., Ramanujam, J., & Li, X. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4853188 · Cited by 1
Facial age estimation through self-paced learning
Li, S., Qiu, S., Liu, H., Zhou, M., & Li, Q. (2017) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip.2017.8305113
“Inspired by [11], the training data can be augmented by hiding patches at different positions, which also gives the data with different difficulty levels.”
Generative Adversarial Networks for Unsupervised Object Co-localizationPreprint
Choe, J., Park, J. H., & Shim, H. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.00236
“Recently, [33,34] introduced that this data augmentation improve localization performance for weakly-supervised object localization.”
BackgroundMethodInfluentialArtifact Detection Maps Learned using Shallow Convolutional Networks
Goodall, T., & Bovik, A. C. (2018) · IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI) · doi:10.1109/ssiai.2018.8470369
“For example, Singh and Yee [32] proposed randomly hiding the most discriminative points during training, to increase generality.”
BackgroundLearning Diversified Features for Object Detection via Multi-region Occlusion Example Generating
Liang, J., Li, Z., & Guo, H. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-16145-3_42
Deep Learning for Action Understanding in Video
Shou, Z. (2019) · doi:10.7916/d8-srkz-t696
“UntrimmedNet [86] and Hide-and-Seek [87] achieve the state-of-the-art performances and carry out the localization in a similar manner.”
BackgroundWeakly Supervised Localization Using Background ImagesPreprint
Kou, Z., Zhao, W., Cui, G., & Wang, S. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.03619
Adaptive Adversarial Example Generating Network for Object Detection
Li, Z., Liang, J., & Guo, H. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-981-13-9406-5_66
Weakly Supervised Object Localization with Inter-Intra Regulated CAMs.Preprint
Cui, G., Kou, Z., Wang, S., Zhao, W., & Xu, C. (2019) · arXiv (Cornell University)
Unified Multifaceted Feature Learning for Person Re-IdentificationPreprint
Cheng, Y., Pang, G., Bai, X. P., & Shen, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1911.08651
“2) Batch-constant Erasing (BcE): Different from the random erasing or other deterministic erasing methods [34], [35] that erase a small rectangle area only and the erasing area is randomly chosen per image, our batch-constant erasing method erases a striping part of the image and the erased part is…”
Channel Attention with Embedding Gaussian Process: A Probabilistic Methodology.Preprint
Xie, J., Chang, D., Ma, Z., Zhang, G., & Guo, J. (2020) · arXiv (Cornell University)
Wang, J., Zhang, R., Wei, X., Li, X., Yu, M., Yu, J., Zhu, J., Gao, J., Shen, H., Liu, Z., et al. (2019)
Weakly Supervised Object Localization Using Self-Paced Pyramid Adversarial Learning
Pan, F.-C., Bian, B., Wang, B., Yang, Y., & Ju, X. (2020) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9207099
“[19] propose Hide-andSeek (HaS) to divide the input image into grid-like patches and randomly selects the patches to erase.”
BackgroundBircanoğlu, C. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.03815
“CAM is a relatively old algorithm and was used in multiple studies [25], [26] as it is or with some variations [27], [28] for multiple purposes such as obtaining saliency maps, object detection [29], instance, and semantic segmentation [30].”
MethodHuang, X., Ma, L., & Tian, J. (2020) · IEEE Region 10 Conference (TENCON) · doi:10.1109/tencon50793.2020.9293843
“To alleviate this, both Hide-and-Seek [7] and Erase-and-collect [8] did gain action areas gradually by iterating hiding object areas in training, and located actions more completely, though when to stop an iteration remains an open question.”
BackgroundCategory‐wise feature extractor based on ADL method for weak‐supervised object localisation
Hu, Y., Zhu, D., Ai, X., & Xu, Y. (2020) · IET Image Processing · doi:10.1049/iet-ipr.2020.0640
Dual Normalization Multitasking for Audio-Visual Sounding Object LocalizationPreprint
Nishikawa, T., Shimada, D., & Yokono, J. J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.00180 · Cited by 1
“CAM and attention, as WSOL, has been criticized for highlighting mainly discriminative regions, and technics such as data augmentations and architectural solutions have been proposed to extract the entire object[35, 38, 39, 12].”
BackgroundGPCA: A Probabilistic Framework for Gaussian Process Embedded Channel AttentionPreprint
Xie, J., Chang, D., Ma, Z., Zhang, G., & Guo, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.04575
“It is mainly motivated for fine-grained visual classification (FGVC) tasks, while previous researches [50], [51], [61], [62], [63], [64], [65] commonly worked on it for WSOL.”
MethodFeature Matching Network for Weakly-Supervised Temporal Action Localization
Dou, P., Zhou, W., Liao, Z., & Hu, H. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88013-2_38
Super-Class Mixup for Adjusting Training Data
Fujii, S., Okamoto, N., Seo, T., Hirakawa, T., Yamashita, T., & Fujiyoshi, H. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-02444-3_4
“And, augmentation based on mask processing is also developed [4, 19, 1, 11].”
BackgroundMethodAttention-based Dual-Branches Localization Network for Weakly Supervised Object Localization
Hui, W., Tan, C., & Gu, G. (2021) · ACM International Conference on Multimedia in Asia (MMAsia) · doi:10.1145/3469877.3490568
“Hide-and-Seek(Has) [14] randomly hid patches in a training image to force the network to seek other relevant part.”
BackgroundWan, X., Liu, L., Gao, Z., & Zhang, X. (2022) · Lecture notes in electrical engineering · doi:10.1007/978-981-19-0572-8_16
Yun, H., Nakamura, E., Kageyama, Y., Ishizawa, C., Kato, N., Igarashi, K., & Kawamoto, K. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4128803
Action Recognition in Real-World Videos
Sultani, W., Arshad, Q. A., & Chen, C. (2021) · Springer eBooks · doi:10.1007/978-3-030-63416-2_846
Learning from limited labelled data: contributions to weak, few-shot, and unsupervised learning
Rodríguez, J. J. S. (2022) · RiuNet (Universitat Politècnica de València) · doi:10.4995/thesis/10251/190633
An Image Segmentation Based on a Piecewise Smooth Model in Deep Learning
Choi, H.-T., Derbel, B., & Hong, B. (2022) · International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) · doi:10.1109/itc-cscc55581.2022.9894902
“Instead, [8] use randomly generated mask which cover the input image.”
BackgroundMethodExploring Pixel Alignment on Shallow Feature for Weakly Supervised Object Localization
Cao, X., Yang, M., & Sun, G. (2022) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892121
“In particular, Data enhancement [6]–[8], [15]–[17] attempts to encourage the network to learn from incomplete data and to be less dependent on discriminative regions.”
BackgroundAn Improved Person Re-Identification Method based on AlignedReID ++ algorithm
Zhu, X., Dong, X., Nie, H., & Cen, Y. (2022) · International Conference on Systems and Informatics (ICSAI) · doi:10.1109/icsai57119.2022.10005320
The Robust Weight Conversion Learning for Classification of Occlusion Images
Kim, J.-H., Ryu, J.-K., & Park, S. (2023) · The Journal of Korea Robotics Society · doi:10.7746/jkros.2023.18.1.122
An Approach for Fusing Two Training-Datasets with Partially Overlapping Classes
Niemeijer, J., Battistella, F., Srinivas, G., & Leich, A. (2023) · IEEE International Conference on Semantic Computing (ICSC) · doi:10.1109/icsc56153.2023.00017
Lan, X., Mei, J., Lin, R., Chen, J., & Zhang, Y. (2023) · Research Square · doi:10.21203/rs.3.rs-2695448/v1
Data Augmentation Algorithm Based on Local Dynamic Transformation
Zhu, H., Zhang, J., Hu, L., Li, C., Shan, M., & Duan, Q. (2022) · International Conference on Artificial Intelligence, Information Processing and Cloud Computing (AIIPCC) · doi:10.1109/aiipcc57291.2022.00090
Diagnosis of Continuous Dynamical Systems Using Structural Sensitivity Functions
Shalobanova, N. P., & Shalobanov, S. S. (2023) · International Conference on Industrial Engineering, Applications and Manufacturing (ICIEAM) · doi:10.1109/icieam57311.2023.10139248
Rough Target Region Extraction with Background Learning
Nakamura, R., Ueda, Y., Tanaka, M., & Fujiki, J. (2023) · Communications in computer and information science · doi:10.1007/978-981-99-4914-4_3
Weakly Supervised Object Localization Using Class-Specific Foreground Maps
Park, S., Lee, S., & Kang, B. (2023) · Journal of Broadcast Engineering · doi:10.5909/jbe.2023.28.4.391
Liu, P., Wang, C., Wang, C., Qin, J., & Wang, J. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4567047
Exploring Information Bottleneck for Weakly Supervised Semantic Segmentation
Qin, J., Lyu, Y., & Wang, X. (2023) · Frontiers in artificial intelligence and applications · doi:10.3233/faia230480
Module of Axis-based Nexus Attention for weakly supervised object localization
Sohn, J., Jeon, E., Jung, W., Kang, E., & Suk, H. (2023) · Scientific Reports · doi:10.1038/s41598-023-45796-8
Weakly Supervised Temporal Action Localization Through Segment Contrastive Learning
Jiang, Z., & Li, Y. (2023) · Communications in computer and information science · doi:10.1007/978-981-99-8141-0_18
CyCSNet: Learning Cycle-Consistency of Semantics for Weakly-Supervised Semantic Segmentation
Duan, Z., Yu, X., & Ding, Y. (2023) · IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences · doi:10.1587/transfun.2023eap1072
Siamese Representation Learning for Weakly Supervised Object Localization
Li, X., Qiu, F., & Wang, B. (2023) · International Conference on Computing, Communication, Perception and Quantum Technology (CCPQT) · doi:10.1109/ccpqt60491.2023.00079
“[5] 74.50 AlexNet-BeamSearch [22] 77.10 VGG16-CAM [7] 78.40 VGG16-BeamSearch [22] 79.70 VGG16-SP [23] 82.90 ProNet [2] 77.70 VGG16-mLDFR [2] 87.01 VGG16-SRLNet 89.25 Tolerance=0 VGG16-SRLNet 87.68 V.”
InfluentialDual-Polarity ViT for Fine-Grained Visual Categorization
Guo, T., Cheng, L., Wang, C., Chen, R., Zhang, H., & Lan, L. (2023) · International Conference on Multimedia Communication Technologies (ICMCT) · doi:10.1109/icmct60483.2023.00016
“Backbone Accuracy ADL(0.25) [35] ResNet-50 40.8% ADL(0.5) [35] ResNet-50 43.8% HaS [36] ResNet-50 48.3% VGG-16 [37] VGG-16 50.8% BYOL ResNet-50 52.92% Cutout(8) [38] ResNet-50 54.6% Cutout(16) [38] ResNet-50 4) Comparison on ultra-fine-grained dataset: Prior works focus on FGVC benchmarks containing…”
InfluentialWeakly-Supervised Temporal Action Localization Based on Attention Regularization
Yuan, R., & Zhang, L. (2023) · International Conference on Communication, Image and Signal Processing (CCISP) · doi:10.1109/ccisp59915.2023.10355856
“Singh et al. proposed the Hide-and-Seek algorithm, which is one of the earliest weakly-supervised TAL methods [1].”
MethodDetection of Defects in Bottled Liquor Using Deep Learning
Samonte, M. J. C., Dong, T., Wu, M., & Wang, Y. (2023) · World Symposium on Software Engineering (WSSE) · doi:10.1145/3631991.3632039
“In the field of computer vision, the methods of data enhancement can be summarized into three aspects, namely Spatial Transformation [11], Color Distortion [12], and Information Dropping [13-15].”
MethodClassification of Chip-Level Defect Types in Wafer Bin Maps Using Only Wafer-Level Labels
Lee, H., Kim, H., & Kim, H. (2024) · Journal of Manufacturing Science and Engineering · doi:10.1115/1.4065226
Small Object Matters in Weakly Supervised Object LocalizationPreprint
Hwang, D., Oh, S. J., & Choe, J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4794778
Xia, H., Zhan, Y., Liu, H., & Ren, X. (2024) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.2300024
“Therefore, some re‐ searchers have studied the erasing mechanism on this basis, such as Hide-and-Seek (Singh and Lee, 2017) and adversarial seeded sequence growing (ASSG) (Zhang et al., 2019).”
ALMA: Adjustable Location and Multi-Angle Attention for Fine-Grained Visual Classification
Ding, B., Xu, X., Bao, X., Yan, N., & Zhang, R. (2024) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd61410.2024.10580689
“A random space image enhancement method [5] was proposed to randomly generate rectangular regions to mask the image during training, thereby improving the robustness of the model.”
MethodDiscovering an inference recipe for weakly-supervised object localization
Lee, S., Mun, C., Uh, Y., Choe, J., & Byun, H. (2024) · Pattern Recognition · doi:10.1016/j.patcog.2024.110838
Building Detection and Segmentation Method in Aerial Photograph Based on Mask R-CNN
Zhang, C., Lin, Y., & Zhu, H. (2024) · International Conference on Computer Communication and Artificial Intelligence (CCAI) · doi:10.1109/ccai61966.2024.10603137
“Commonly employed data augmentation techniques include image rotation, flipping, scaling, and panning, while more sophisticated methods involve introducing noise perturbations into the image or selectively masking regions in the input image [15-16].”
MethodA mixed sample data augmentation method based on sparse adversarial attackPreprint
Xu, X., Zhang, X., Wang, Y., Liu, L., & Xu, E. (2024) · Research Square · doi:10.21203/rs.3.rs-4698850/v1
ESC: Explaining the Predictions of Deep Neural Networks via Concept Similarity
Gao, L.-J., Wang, S., Gu, W., Sun, Z., Jin, X., & Zhang, Y. (2024) · IEEE International Joint Conference on Neural Network · doi:10.1109/ijcnn60899.2024.10650575
“Some studies leverage changes at the input layer[30, 31] and intermediate layers[32] for supervised localization.”
BackgroundA Whale Falls, All Thrive: Mitigating Attention Gap to Improve Adversarial Transferability
Wan, Q., Chen, B., Peng, A., & Zeng, H. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-78312-8_23
Duhan, S., Gulia, P., & Gill, N. S. (2024) · Innovations in Sustainable Technologies and Computing · doi:10.1007/978-981-97-7831-7_4
Rethinking erasing strategy on weakly supervised object localization
Fan, Y., Wei, S., Tan, C., Chen, X., Yang, D., & Zhao, Y. (2025) · Signal Processing Image Communication · doi:10.1016/j.image.2025.117280
Dual-branch contrastive learning for weakly supervised object localization
Guo, Z., Li, D., Du, Z., & Seng, B. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06514-1
“ACoL [9] and ADL [10] adopt adversarial erasing and utilize adver-sarially trained classifiers to reconstruct missing regions.”
A Realistic Protocol for Evaluation of Weakly Supervised Object Localization
Murtaza, S., Belharbi, S., Pedersoli, M., & Granger, É. (2025) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv61041.2025.00524 · Cited by 1
“Although this protocol allows us to avoid using the test set and to perform unbiased modal evaluations, it is somewhat unrealistic since annotated bboxes are not available in practice.”
InfluentialUsing Off-the-Shelf Deep Neural Networks for Position-Based Visual Servoing
Tschesche, M., Hofmann, T., Ferrein, A., & Lakemeyer, G. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-85859-8_3
“Additionally, the dataset was augmented by applying cropping and padding, hide and seek [33], and CutMix [39], using random images from the COCO dataset [22] as replacement.”
MethodConvolutional neural network for data augmentation
Kim, R., & White, E. (2024) · World Journal of Advanced Engineering Technology and Sciences · doi:10.30574/wjaets.2024.13.2.0528 · Cited by 6
Intra-Class Patch Swap for Self-Distillation
Choi, H., Jeon, E. S., Shukla, A., & Turaga, P. K. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.130408
GraphLOMO: LOcating multiple objects without visual annotations
To, A., Davis, J. G., & Nguyen, H. D. (2025) · Machine Vision and Applications · doi:10.1007/s00138-025-01707-7
“Previous solutions to alleviate the problems include architectural modifi - cations [8, 11, 12] and data augmentation [9, 10], although these methods still require implicit full supervision at some point for hyper-parameter and model selection.”
Li, Y., Ma, M., Yao, C., Pei, Z., & Wu, J. (2025) · Neurocomputing · doi:10.1016/j.neucom.2025.131304
ShapeAug++: More Realistic Shape Augmentation for Event Data
Bendig, K., Schuster, R., & Stricker, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06007-5_1
Cross-slide augmentation for whole slide image classification based on class activation map
Chen, Y., Wu, H., Huang, Z., Zhang, Z., & Cai, H. (2025) · Scientific Reports · doi:10.1038/s41598-025-26113-x
“The emergence of CAM has led to a consistent approach in these weakly supervised tasks, proposing CAM-based weakly supervised frameworks 33–36 .”
Graph-Based Unsupervised Temporal Segmentation of Diving Actions
Badatya, B. K., Baghel, V., & Hegde, R. S. (2025) · IEEE International Workshop on Sport, Technology and Research (STAR) · doi:10.1109/star66750.2025.11264778
Metadata
- Authors
- Singh, K. K., & Lee, Y. J.
- Affiliations
- Intel (United States) (US)
- University of California, Davis (US)
- Venue
- 2017 IEEE International Conference on Computer Vision (ICCV)
- Publisher
- IEEE
- Published
- 2017-10
- Pages
- 3544-3553
- OpenAlex
- W2606520630
- Cited by
- 724
- References
- 57
- Citations by year
- 2026: 19 · 2025: 49 · 2024: 74 · 2023: 121 · 2022: 129 · 2021: 137 · 2020: 98 · 2019: 70 · 2018: 25 · 2017: 2
- FWCI
- 15.59
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Nvidia
- Amazon Web Services
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Image and Video Retrieval Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Visual Attention and Saliency Detection
- Keywords
- weakly supervised object localization, data augmentation, image classification, video action recognition
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
