Ensemble of exemplar-SVMs for object detection and beyond
Open accessThe DOI matches its Crossref record; 23 of the 25 references it cites were found in a registry or an index.
Abstract
This paper proposes a conceptually simple but surprisingly powerful method which combines the effectiveness of a discriminative object detector with the explicit correspondence offered by a nearest-neighbor approach. The method is based on training a separate linear SVM classifier for every exemplar in the training set. Each of these Exemplar-SVMs is thus defined by a single positive instance and millions of negatives. While each detector is quite specific to its exemplar, we empirically observe that an ensemble of such Exemplar-SVMs offers surprisingly good generalization. Our performance on the PASCAL VOC detection task is on par with the much more complex latent part-based model of Felzenszwalb et al., at only a modest computational cost increase. But the central benefit of our approach is that it creates an explicit association between each detection and a single training exemplar. Because most detections show good alignment to their associated exemplar, it is possible to transfer any available exemplar meta-data (segmentation, geometric structure, 3D model, etc.) directly onto the detections, which can then be used as part of overall scene understanding.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures


![Figure 3 Comparison. Given a bicycle training sample from PASCAL (represented with a HOG weight vector w ), we show the top 6 matches from the PASCAL test-set using three methods. Row 1: naive nearest neighbor (using raw normalized HOG). Row 2: Trained Exemplar-SVM (notice how w focuses on bike-specific edges). Row 3: Learned distance function - an Exemplar-SVM but trained in the 'distance-to-exemplar' vector space, with the exemplar being placed at the origin (loosely corresponding to [10, 14]).](https://figures.cochat.ai/paper-figures/1cc4b956c3534a219577db00fbbb8da4/ensemble-of-exemplar-svms-for-object-detection-and-beyond-figure-3.png)







Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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
“In the case of sliding-window object detectors, such as the deformable parts model (DPM) [12], this imbalance may be as extreme as 100,000 background examples to every one object.The recent trend towards object-proposal-based detectors [15,32] mitigates this issue to an extent, but the imbalance ratio may still be high (e.g., 70:1).”
From Training Region-Based Object Detectors with Online Hard Example Mining (2016) · cites this paper as [32]
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
LIBSVM: A library for support vector machines
Chang, C.-C., & Lin, C.-J. (2011) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/1961189.1961199 · Cited by 40,918 · cited 1×
“Using the procedure above, we train an ensemble of Exemplar-SVM, one for each positive instance in the training set.”
MethodHistograms of Oriented Gradients for Human Detection
Dalal, N., & Triggs, B. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.177 · Cited by 33,584 · cited 2×
“Yet today, a number of methods can do just that with reasonable accuracy.”
BackgroundThe pascal visual object classes (VOC) challenge
Everingham, M., Van Gool, L., Williams, C. K. I., Winn, J., & Zisserman, A. (2010) · International Journal of Computer Vision · doi:10.1007/s11263-009-0275-4 · Cited by 19,599 · cited 2×
“For our experiments, we use a single source of exemplars: the PASCAL VOC 2007 dataset [8] – a popular dataset used to benchmark object detection algorithms.”
BackgroundMethodLibsvm : A library for support vector machines
Ferrari, V. (2008) · Medical Entomology and Zoology · Cited by 10,122
Object Detection with Discriminatively Trained Part-Based Models
Felzenszwalb, P. F., Girshick, R. B., McAllester, D., & Ramanan, D. (2010) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2009.167 · Cited by 9,853 · cited 4×
“For this, we present results on a set of tasks including segmentation, qualitative-geometry estimation, 3D model transfer, and related object priming.”
BackgroundMethodInfluentialProbabilistic Outputs for Support vector Machines and Comparisons to Regularized Likelihood Methods
Platt, J. C. (1999) · Cited by 6,384
Shape matching and object recognition using shape contexts
Belongie, S., Malik, J., & Puzicha, J. (2002) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/34.993558 · Cited by 6,306 · cited 1×
“Because most detections show good alignment to their associated exemplar, it is possible to transfer any available exemplar meta-data (segmentation, geometric structure, 3D model, etc.) directly onto the detections, which can then be used as part of overall scene understanding.”
BackgroundEstimating the Support of a High-Dimensional Distribution
Schölkopf, B., Platt, J. C., Shawe-Taylor, J., Smola, A. J., & Williamson, R. C. (2001) · Neural Computation · doi:10.1162/089976601750264965 · Cited by 6,168
SUN database: Large-scale scene recognition from abbey to zoo
Xiao, J., Hays, J., Ehinger, K. A., Oliva, A., & Torralba, A. (2010) · 2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2010.5539970 · Cited by 3,283 · cited 1×
“Because most detections show good alignment to their associated exemplar, it is possible to transfer any available exemplar meta-data (segmentation, geometric structure, 3D model, etc.) directly onto the detections, which can then be used as part of overall scene understanding.”
Background80 Million Tiny Images: A Large Data Set for Nonparametric Object and Scene Recognition
Torralba, A., Fergus, R., & Freeman, W. T. (2008) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2008.128 · Cited by 1,921
Research works printed but not linked: 1
- A. Frome and J. Malik. Image retrieval and recognition using local distance functions. NIPS, 2006. 2, 3, 4
Cited by
Items 501–600 of 873
Olejniczak, M., & Kraft, M. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-59063-9_49 · Cited by 8
Patch-wise Contrastive Style Learning for Instagram Filter Removal
Kınlı, F., Özcan, B., & Kıraç, F. (2022) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00073 · Cited by 8
Part level transfer regularization for enhancing exemplar SVMs
Aytar, Y., & Zisserman, A. (2015) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2015.04.004 · Cited by 10
Discriminative and consistent similarities in instance-level Multiple Instance Learning
Rastegari, M., Hajishirzi, H., & Farhadi, A. (2015) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2015.7298674 · Cited by 8
Document Retrieval with Unlimited Vocabulary
Ranjan, V., Harit, G., & Jawahar, C. V. (2015) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2015.104 · Cited by 8
“For searching complex concepts in large databases, SVMs have emerged as the most popular and accurrate solution in the recent past [10].”
BackgroundActivity recognition and prediction with pose based discriminative patch model
Cao, S., Chen, K., & Nevatia, R. (2016) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2016.7477584 · Cited by 8
“proposed mid-level discriminative patches based on Exemplar-SVM [16].”
BackgroundActive Metric Learning for Object Recognition
Ebert, S., Fritz, M., & Schiele, B. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-32717-9_33 · Cited by 8
“for object detection [12] or human pose estimation [15].”
BackgroundEfficient face frontalization in unconstrained images
Mallikarjun, B. R., Chari, V., & Jawahar, C. V. (2015) · National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics (NCVPRIPG) · doi:10.1109/ncvpripg.2015.7490057 · Cited by 10
“In line with our approach, some recent works try to counter these challenging conditions of pose variation by synthesizing pose neutral faces from input images.”
BackgroundImprovement on Deep Features through Various Enhancement Techniques for Vehicles Classification
Kiran, V. K., Dash, S., & Parida, P. (2021) · Sensing and Imaging · doi:10.1007/s11220-021-00363-1 · Cited by 8
“SVM has demonstrated its efficacy in classification problems, namely vehicle classification, facial detection, and object recognition [51, 52].”
BackgroundObject Classification with Adaptable Regions
Bilen, H., Pedersoli, M., Namboodiri, V. P., Tuytelaars, T., & Gool, L. V. (2014) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2014.468 · Cited by 11
“A better strategy is to train an exemplar classifier, such as exemplar SVM [15], for each background region in positive images.”
BackgroundJoint Inference of Objects and Scenes With Efficient Learning of Text-Object-Scene Relations
Wang, B., Lin, D., Xiong, H., & Zheng, Y. F. (2016) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2016.2520087 · Cited by 8
“To evaluate the proposed model in object classification, we adopt the classic SPM as the scene classifier, and test three object detectors: the classic deformable part based model (DPM) [1], the exemplar-SVM (ESVM) [35], and the state-of-the-art RCNN”
MethodInfluentialMultiple Granularity Modeling: A Coarse-to-Fine Framework for Fine-grained Action Analysis
Ni, B., Paramathayalan, V. R., Li, T., & Moulin, P. (2016) · International Journal of Computer Vision · doi:10.1007/s11263-016-0891-8 · Cited by 8
“(7) Note that the detection scores for various models are calibrated according to the method used in Malisiewicz et al. (2011).”
Gradient Networks: Explicit Shape Matching Without Extracting Edges
Hsiao, E., & Hebert, M. (2013) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v27i1.8559 · Cited by 8
“We use the one hundred negative images in CMU KO8, the same parameters as (Malisiewicz and Efros 2011), and three hard negative mining iterations for training.”
MethodInfluentialSemantic Highlight Retrieval and Term Prediction
Sun, M., Zeng, K.-H., Lin, Y.-C., & Farhadi, A. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2666039 · Cited by 8
“1) Calibrating Classifier Confidence: Similar to [56], we calibrate the confidences by fitting a probability distribution [57] on a set of hold-out data.”
MethodA Classification Algorithm for Blind UAV Detection in Wideband RF Systems
Prasad, K. N. R. S. V., & Bhargava, V. K. (2020) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2020-fall49728.2020.9348513 · Cited by 8
Semantic Understanding of Professional Soccer CommentariesPreprint
Hajishirzi, H., Rastegari, M., Farhadi, A., & Hodgins, J. K. (2012) · arXiv (Cornell University) · doi:10.48550/arxiv.1210.4854 · Cited by 8
“Recently in computer vision, Exemplar Support Vector Machines (ESVM) has shown great success in learning what is unique about an image that can distinguish it from all other images [Malisiewicz et al., 2011; Shrivastava et al., 2011].”
BackgroundSoft biometrics in low resolution and low quality CCTV videos
Semertzidis, T., Αξενόπουλος, Α., Karadimos, P., & Daras, P. (2016) · International Conference on Imaging for Crime Detection and Prevention (ICDP) · doi:10.1049/ic.2016.0092 · Cited by 8
“Motivated by [10], our aim is to provide a fast and accurate segmentation of each detected person, to enhance the sampling quality and, subsequently, the overall accuracy of the extracted soft biometric labels.”
BackgroundMethodWhich Strategies Matter for Noisy Label Classification? Insight into Loss and UncertaintyPreprint
Shin, W., Ha, J.-W., Li, S., Cho, Y., Song, H., & Kwon, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.06218 · Cited by 9
“Difficult samples are known to accelerate training, especially for datasets with a small amount of noise [11].”
BackgroundJoint calibration of Ensemble of Exemplar SVMs
Modolo, D., Vezhnevets, A. S., Russakovsky, O., & Ferrari, V. (2015) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2015.7299021 · Cited by 7
Spatiotemporal Feature Enhancement Network for Blur Robust Underwater Object Detection
Zhou, H., Qi, L., Huang, H. Y., Yang, X., & Yang, J. (2024) · IEEE Transactions on Cognitive and Developmental Systems · doi:10.1109/tcds.2024.3386664 · Cited by 8
“Later on, multiple proposals [27], [28], [29] were made to improve the performance of HOG in different object detection tasks.”
BackgroundPredicting an Object Location Using a Global Image Representation
Serrano, J. A. R., & Larlus, D. (2013) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2013.217 · Cited by 8
“In the first strategy, approaches still perform sliding window search and each sub-window is used as a query for which nearest-neighbor similarity is computed [31, 22].”
MethodAccelerated Deformable Part Models on GPUs
Hirabayashi, M., Kato, S., Edahiro, M., Takeda, K., & Mita, S. (2015) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2015.2453962 · Cited by 7
“SVMs are now used in many objectdetection algorithms [29], [5].”
MethodWeighted SVM with classification uncertainty for small training samples
Choi, M.-K., Lee, H.-G., & Lee, S. (2016) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2016.7533199 · Cited by 7
Online learning and detection of faces with low human supervision
Villamizar, M., Sanfeliu, A., & Moreno-Noguer, F. (2018) · The Visual Computer · doi:10.1007/s00371-018-01617-y · Cited by 7
“Machine learning techniques like Support Vector Machines [7,9,35], Boosting [2,31,49,54,56], Random Forests [6,12, 39], and more recently Deep Learning [19,27,40] are commonly used to build classifiers robust to large intra-class variations and other artifacts such as loss of image resolution, rotations, deformations…”
BackgroundMethodBrick Kiln Detection and Localization using Deep Learning Techniques
Tahir, R., Imran, M. S., Minhas, S., Sabahat, N., Ilyas, S. H. W., & Gadi, H. R. (2021) · doi:10.1109/icai52203.2021.9445267 · Cited by 7
“In a study carried out at the Carnegie Melon University [11], the Support Vector Machine (SVM) was combined with the nearest neighbour approach to perform object detection and classification on the PASCAL VOC-2007 dataset.”
MethodDiscriminative tree-based feature mapping
Kobetski, M., & Sullivan, J. (2013) · British Machine Vision Conference (BMVC) · doi:10.5244/c.27.71 · Cited by 7
“Exemplar-SVMs can be seen as an extreme case of this, where each example is associated with one linear decision boundary [13].”
BackgroundData Decomposition and Spatial Mixture Modeling for Part Based Model
Zhang, J., Huang, Y., Huang, K., Wu, Z., & Tan, T. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-37331-2_10 · Cited by 7
“ExModel is an interesting method introduced recently in [28].”
MethodClothing Co-Segmentation Based on HOG Features and E-SVM Classifier
Zhang, J., Liu, L., Huang, D., Fu, X., & Huang, Q. (2016) · International Conference on Digital Home (ICDH) · doi:10.1109/icdh.2016.013 · Cited by 7
GenELM: Generative Extreme Learning Machine feature representation
Zhou, S., Deng, C., Wang, W., Huang, G.-B., & Zhao, B. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.05.098 · Cited by 7
Hierarchical representation learning using spherical k-means for segmentation-free word spotting
Mhiri, M., Abuelwafa, S., Desrosiers, C., & Cheriet, M. (2017) · Pattern Recognition Letters · doi:10.1016/j.patrec.2017.11.010 · Cited by 7
Lahiri, A., Ragireddy, S. C., Biswas, P. K., & Mitra, P. (2019) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2019.00197 · Cited by 7
Real-Time Driver Distraction Detection Using OpenCV and Machine Learning Algorithms
Swathi, V. N. V. L. S., Akhilesh, D., Kumar, G. S., & Vathsala, A. V. (2021) · Smart innovation, systems and technologies · doi:10.1007/978-981-16-0878-0_5 · Cited by 7
Surprising Effectiveness of Few-Image Unsupervised Feature Learning.Preprint
Asano, Y. M., Rupprecht, C., & Vedaldi, A. (2019) · arXiv (Cornell University) · Cited by 8
Du, G., Wang, K., Lian, S., & Zhao, K. (2019) · arXiv (Cornell University) · Cited by 7
MetaInfoNet: Learning Task-Guided Information for Sample ReweightingPreprint
Wei, H., Feng, L., Wang, R., & An, B. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.05273 · Cited by 7
“Sample reweighting algorithms [20, 23, 26] are widely studied for robust learning from biased data.”
An End-to-End deep learning system for writer identification in handwritten Arabic manuscripts
Chammas, M., Makhoul, A., Demerjian, J., & Dannaoui, E. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-17303-8 · Cited by 6
“AnExemplar-SVM(ESVM)withalinearkernelwasimplemented[14, 31],whichresultsanimprovedaccuracyandprecision.”
A hardware accelerated multilevel visual classifier for embedded visual-assist systems
Cotter, M. J., Advani, S., Sampson, J., Irick, K. M., & Narayanan, V. (2014) · doi:10.5555/2691365.2691388 · Cited by 6
Revealing What to Extract from Where, for Object-Centric Content Based Image Retrieval (CBIR)
Gupta, N., Das, S., & Chakraborti, S. (2014) · Indian Conference on Computer Vision, Graphics and Image Processing · doi:10.1145/2683483.2683540 · Cited by 6
“In [21], use of an ensemble of Exemplar-SVMs for object detection has been proposed.”
MethodStatistical and Spatial Consensus Collection for Detector Adaptation
Sangineto, E. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-10578-9_30 · Cited by 6
“Similarly, in [21] the context score is linearly combined using the pre-computed pairwise co-occurrence frequencies of different classifiers, which need to be trained in a supervised manner.”
BackgroundMethodLW-ODF: A Light-Weight Object Detection Framework for Optical Remote Sensing Imagery
Wu, X., Hong, D., Ghamisi, P., Li, W., & Tao, R. (2019) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss.2019.8898673 · Cited by 6
“Method COPD [11] BOW-SVM [15] Exemplar [16] ACF [5] YOLO1 [17] YOLO2 [18] CCF [10] Ours”
Non-parametric scene parsing: Label transfer methods and datasets
Bhowmick, A., Saharia, S., & Hazarika, S. M. (2022) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2022.103418 · Cited by 7
A Review of Space Target Recognition Based on Ensemble Learning
Wang, S., Danpu, Z., Hong, H., & Sun, K. (2025) · Aerospace · doi:10.3390/aerospace12040278 · Cited by 6
Subcategory Clustering with Latent Feature Alignment and Filtering for Object Detection
Ruan, Z., Wang, G., Xue, J., & Lin, X. (2014) · IEEE Signal Processing Letters · doi:10.1109/lsp.2014.2349940 · Cited by 6
“The second direction is based on the exemplar SVM [8].”
BackgroundIntegrated instance- and class-based generative modeling for text classification
Puurula, A., & Myaeng, S.-H. (2013) · Australasian Document Computing Symposium · doi:10.1145/2537734.2537751 · Cited by 6
“Instance- and class-based learning have also been combined to design hybrid methods that outperform both types of learning on their own [25, 5, 18].”
MethodProperties of Datasets Predict the Performance of Classifiers
Aghazadeh, O., & Carlsson, S. (2013) · British Machine Vision Conference (BMVC) · doi:10.5244/c.27.44 · Cited by 6
“However, the discriminative feature selection embedded in the class specific similarity measure - through the use of exemplar SVMs - already knows what does not belong to a class, locally in the space.”
MethodFast Cross-Scenario Clothing Retrieval Based on Indexing Deep Features
Li, Z., Li, Y., Gao, Y., & Liu, Y. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-48890-5_11 · Cited by 6
Video-Specific SVMs for Colonoscopy Image Classification
Manivannan, S., Wang, R., Trujillo, M., Hoyos, J. A., & Trucco, E. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-13410-9_2 · Cited by 6
“As in [14], we use the Platt calibration method [15] to normalize individual classifiers such that their outputs are more directly comparable.”
BackgroundMethodInfluentialGenre-based image classification using ensemble learning for online flyers
Pourashraf, P., Tomuro, N., & Apostolova, E. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2197138 · Cited by 7
“Recent works include [3] and [4].”
Store classification using Text-Exemplar-Similarity and Hypotheses-Weighted-CNN
Huang, C., Li, H., Li, W., Wu, Q., & Xu, L. (2017) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2017.01.011 · Cited by 6
“ExemplarSVM [36] is employed to train the detector of each candidate region.”
MethodInstance-based object recognition in 3D point clouds using discriminative shape primitives
Zhang, J., & Sun, J. (2017) · Machine Vision and Applications · doi:10.1007/s00138-017-0885-8 · Cited by 6
3D Object Classification Using 2D Perspectives of Point Clouds
Jansen, L., Liebrecht, N., Soltaninejad, S., & Basu, A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-54407-2_38 · Cited by 6
Site Detection for Autonomous Soft-Landing on Asteroids Using Deep Learning
Ravani, K., Mathavaraj, S., & Padhi, R. (2021) · Transactions of Indian National Academy of Engineering · doi:10.1007/s41403-021-00207-0 · Cited by 6
“Examples of such algorithms include histogram of oriented gradient-based techniques (Malisiewicz et al. 2011).”
BackgroundSupervised Adaptive-RPN Network for Object Detection in Remote Sensing Images
Tang, X., Zhang, H., Ma, J., Zhang, X., & Jiao, L. (2020) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss39084.2020.9323998 · Cited by 6
“To further study our model, we compare it with several methods, including Exemplar-SVMs [11], the collection of part detectors (COPD) [10], rotation-invariant CNN (RICNN) [12], HyperNet [13], and the rotation-insensitive and context-augmented (RICA) [14] method.”
MethodCustomer Analysis via Video Analytics: Customer Detection with Multiple Cues
(2017) · Acta Polytechnica Hungarica · doi:10.12700/aph.14.3.2017.3.11 · Cited by 6
“[27] predicts a class based on combined predictions from multiple linear Support Vector Machines (SVMs), each trained on only one example.”
BackgroundClustering based ensemble correlation tracking
Zhu, G., Wang, J., & Lu, H. (2016) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2016.05.006 · Cited by 6
Car Segmentation and Pose Estimation using 3D Object ModelsPreprint
Mahendran, S., & Vidal, R. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1512.06790 · Cited by 6
“The only difference from the standard Structural SVM problem is that we need non-negative weights.”
MethodUnifying Heterogeneous Classifiers with DistillationPreprint
Vongkulbhisal, J., Vinayavekhin, P., & Visentini-Scarzanella, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.06062 · Cited by 9
“, multi-class classification [15], object detection [34, 27], tracking [1], etc.”
BackgroundZeng, K.-H., Lin, Y.-C., Farhadi, A., & Sun, M. (2016) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2016.7532982 · Cited by 5
“We further calibrate the classifiers similar to [11], since classifiers are trained separately and their virlets representation x are not comparable (see supplementary material for details).”
MethodMeta-optimized Angular Margin Contrastive Framework for Video-Language Representation Learning
Nguyen, T. T., Bin, Y., Wu, X., Dong, X., Hu, Z., Le, K., Nguyen, C.-D. T., Ng, S.-K., & Tuan, L. A. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72989-8_5 · Cited by 5
Exemplar-based logo and trademark recognition
Farajzadeh, N. (2015) · Machine Vision and Applications · doi:10.1007/s00138-015-0695-9 · Cited by 5
“For example, if we do not have many cars in our dataset, the best we may do is to associate the observed novel car to another car [1], i.”
BackgroundMethodInfluentialLearning Detectors Quickly with Stationary Statistics
Valmadre, J., Sridharan, S., & Lucey, S. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-16865-4_7 · Cited by 7
“Examples include adaptive tracking [1, 2], object detection with a large number of classes [3, 4], algorithms for discovering discriminable clusters [5–7] and exemplar-based methods which train a detector per example [8].”
BackgroundExemplar-based linear discriminant analysis for robust object tracking
Gao, C., Chen, F., Yu, J.-G., Huang, R., & Sang, N. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025077 · Cited by 5
“Malisiewicz et al . proposed exemplar-based support vector machine (SVM) algorithm for object detection in [12].”
BackgroundMethodTransfer learning via attributes for improved on-the-fly classification
Kulkarni, P., Sharma, G., Zepeda, J., & Chevallier, L. (2014) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2014.6836097 · Cited by 5
Describing and learning of related parts based on latent structural model in big data
Liu, L., Bai, X. P., Zhang, H., Zhou, J. W., & Tang, W. (2015) · Neurocomputing · doi:10.1016/j.neucom.2014.12.120 · Cited by 5
“Here we use the exemplar-SVM method proposed in [44].”
Interactive Image Retrieval for Biodiversity Research
Freytag, A., Schadt, A., & Denzler, J. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24947-6_11 · Cited by 6
“Unfortunately, training of Exemplar-SVMs involves computationally expensive hard negative mining, In contrast, Exemplar-LDA models can be trained highly efficiently since the majority of computations is done only once in an offline stage.”
MethodMultiple Instance Models Regression for Robust Visual Tracking
Zha, Y., Zhang, Y., Ku, T., Huang, H., Huang, W., & Zhang, P. (2020) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2020.2996621 · Cited by 5
“The optimization of the proposed ensembled model can be regarded as a prime problem [21], which is usually transferred to a corresponding dual formulation [22] to solve.”
BackgroundMethodInfluentialMeta-Learning for Decoding Neural Activity Data With Noisy Labels
Xu, D., & Chen, R. (2022) · Frontiers in Computational Neuroscience · doi:10.3389/fncom.2022.913617 · Cited by 5
“The sample reweighting strategy has been well studied inclduing boosting (Freund and Schapire, 1997), hard sample mining (Malisiewicz et al., 2011), and focal loss (Lin et al., 2017).”
BackgroundVisual Geo-localization of Non-photographic Depictions via 2D–3D Alignment
Aubry, M., Russell, B. C., & Šivic, J. (2016) · Advances in computer vision and pattern recognition · doi:10.1007/978-3-319-25781-5_14 · Cited by 5
“Note that a similar idea has been used in learning per-exemplar distances [20] or per-exemplar support vector machine (SVM) classifiers [36] for object recognition and cross-domain image retrieval [47].”
BackgroundMethodA poselet based key frame searching approach in sports training videos
Wu, L., Zhang, J., & Yan, F. (2012) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · Cited by 5
Pedestrian Counting for a Large Scene Using a GigaPan Panorama and Exemplar-SVMs
Zhang, S., Yu, S., Ma, Q., Shang, P., Gui, P., Wang, J., & Feng, T. (2013) · International Conference on Computational Intelligence and Security (CIS) · doi:10.1109/cis.2013.55 · Cited by 5
“[10] and showed excellent performance in pedestrian detection.”
BackgroundMethodSVM with OpenCL: High performance implementation of support vector machines on heterogeneous systems
Peters, E., & Savakis, A. E. (2015) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2015.7351622 · Cited by 5
“Index Terms— Support Vector Machines, OpenCL, Heterogeneous Computing, GPU Acceleration”
BackgroundDiscriminative learning based visual servoing across object instances
Pandya, H., Krishna, K. M., & Jawahar, C. V. (2016) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2016.7487523 · Cited by 5
Retrieving indoor objects: 2D-3D alignment using single image and interactive ROI-based refinement
Liu, F., Wang, S., Ding, D., Yuan, Q., Yao, Z., Pan, Z., & Li, H. (2017) · Computers & Graphics · doi:10.1016/j.cag.2017.07.029 · Cited by 5
“Otherwise, we would have to try all the object categories and calibrate matching scores among different categories as is done with Exemplar-SVM [39] because matching scores are computed based on discriminative features”
BackgroundMethodInfluentialCategorizing scenes by exploring scene part information without constructing explicit models
Bai, S., & Tang, H. (2017) · Neurocomputing · doi:10.1016/j.neucom.2017.12.003 · Cited by 6
Robust Aircraft Detection with a Simple and Efficient Model
Zhong, J., Leí, T., Yao, G., & Jiang, P. (2018) · Information · doi:10.3390/info9040074 · Cited by 5
“The methods in [4,5] were adopted for comparison.”
BackgroundMethodInfluentialWeighted-learning-instance-based retrieval model using instance distance
Wu, H., Li, Y., Xiong, J., Bi, X., Zhang, L., Bie, R., & Guo, J. (2018) · Machine Vision and Applications · doi:10.1007/s00138-018-0988-x · Cited by 6
“Moreover, the SVM model is sufficiently powerful to support other techniques, such as image classification [7], image annotation [8], and object detection [9, 10].”
MethodA Mixed Method for Feature Extraction Based on Resonance Filtering
Zhang, X., Lu, W., Ding, Y., Song, Y., & Xia, J. (2022) · Intelligent Automation & Soft Computing · doi:10.32604/iasc.2023.027219 · Cited by 5
An automatic methodology to measure drivers’ behavior in public transport
Catalan, H. F., Löbel, H., & Herrera, J. C. (2022) · Journal of Intelligent Transportation Systems · doi:10.1080/15472450.2022.2129019 · Cited by 5
“This approach is adapted from Malisiewicz et al. (2011), where a highly imbalanced set is used in the context of image classification.”
MethodSiamSampler: Video-Guided Sampling for Siamese Visual Tracking
Li, P., Chen, B., Bai, L., Qiao, L., Li, B., & Ouyang, W. (2022) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2022.3214480 · Cited by 6
“The idea of data sampling has been well studied in the literature [15], [20], [26], [37], [39], [44].”
BackgroundObservations on K-Image Expansion of Image-Mixing Augmentation
Jeong, J., Cha, S., Choi, J., Yun, S., Moon, T., & Yoo, Y. (2023) · IEEE Access · doi:10.1109/access.2023.3243108 · Cited by 8
“An intermediate variable v is then selected using SBP sampled from φ (see Line 6, 7, 12, and 13).”
Karmaker, D., Schiffner, I., Strydom, R., & Srinivasan, M. V. (2016) · International Conference on Control, Automation, Robotics and Vision (ICARCV) · doi:10.1109/icarcv.2016.7838650 · Cited by 5
Subcategory-Aware Object Detection
Yu, X., Yang, J., Lin, Z., Wang, J., Wang, T., & Huang, T. S. (2014) · IEEE Signal Processing Letters · doi:10.1109/lsp.2014.2299571 · Cited by 5
“Exemplar-SVMs [2] is a extreme case of multiple components model, which treated every positive sample in the training dataset as a subspace, and trained ensemble of large number of detectors.”
BackgroundHinge-Minimax Learner for the Ensemble of Hyperplanes
Raviv, D., Hazan, T., & Osadchy, M. (2018) · Journal of Machine Learning Research · Cited by 5
Combining UAV-imagery and machine learning for wildlife conservation
Nicolas, R. (2016) · Infoscience (Ecole Polytechnique Fédérale de Lausanne) · Cited by 5
EXEMPLAR-BASED LINEAR DISCRIMINANT ANALYSIS FOR ROBUST OBJECT TRACKINGPreprint
Gao, C., Chen, F., Yu, J.-G., Huang, R., & Sang, N. (2016) · Cited by 5
On Optimizing Ensemble Models using Column Generation
Aziz, V., Wu, O., Nowak, I., Hendrix, E. M. T., & Kronqvist, J. (2024) · Journal of Optimization Theory and Applications · doi:10.1007/s10957-024-02391-9 · Cited by 4
“There are plenty of real-world applications of ensemble learning such as object detection, object recognition [1, 16, 17, 26], fault detection and diagnosis [11, 15, 25], semantic segmentation, edge detection [2, 10, 11], and for black box optimization [23].”
Hayashi, M., Oshima, K., Tanabiki, M., & Aoki, Y. (2015) · IPSJ Transactions on Computer Vision and Applications · doi:10.2197/ipsjtcva.7.18 · Cited by 4
“Exemplar-SVM [23] is an object detection method using perexemplar detectors.”
MethodKernelized pyramid nearest-neighbor search for object categorization
Cheng, H., Yu, R., Liu, Z., Yang, L., & Chen, X. (2014) · Machine Vision and Applications · doi:10.1007/s00138-014-0608-3 · Cited by 4
“Inspired by the idea of calculating posterior probability of SVMs, we use the Platt’s algorithm [26,31] to calculate the posterior probability of the output of each pyramid SVMs.”
MethodAdaptive Exponential Smoothing for Online Filtering of Pixel Prediction Maps
Dang, K., Yang, J., & Yuan, J. (2015) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.367 · Cited by 4
“We can train a classifier such as linear SVM or logistic regression to further perform score calibration [23, 31].”
MethodClustered Exemplar-SVM: Discovering sub-categories for visual recognition
Shapovalova, N., & Mori, G. (2015) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2015.7350766 · Cited by 4
“[4], where each exemplar forms a subcategory of its own, and a general classifier is replaced by an Ensemble of Exemplar SVMs.”
BackgroundMethodInfluentialDiscriminative local binary pattern
Kobayashi, T. (2016) · Machine Vision and Applications · doi:10.1007/s00138-016-0780-8 · Cited by 4
“It is superior to 1-NN in that individual classifiers contain discriminative information endowed by SVM [13, 21].”
BackgroundMethod3D object recognition from large-scale point clouds with global descriptor and sliding window
Gunji, N., Niigaki, H., Tsutsuguchi, K., Kurozumi, T., & Kinebuchi, T. (2016) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2016.7899720 · Cited by 6
“In order to cope with pose variances of target objects, their method requires training of many Exemplar-SVMs, each of which accounts for the same object in a different pose.”
MethodInfluentialSuperpixel Correspondence for Non-parametric Scene Parsing of Natural Images
Naosekpam, V., Bhowmick, A., & Hazarika, S. M. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-34869-4_67 · Cited by 4
Semantic Document Layout Analysis of Handwritten Manuscripts
Jaha, E. S. (2023) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2023.036169 · Cited by 4
Object Detection Algorithms - A Review
Letavay, M., Bažant, M., & Tuček, P. (2023) · International Conference on Control, Artificial Intelligence, Robotics & Optimization (ICCAIRO) · doi:10.1109/iccairo58903.2023.00014 · Cited by 4
“It has been particularly successful in pedestrian detection, where the HOG features help distinguish human-like shapes from the background, [4], [5], [6].”
BackgroundInteractive Known-Item Search Using Semantic Textual and Colour Modalities
Zhang, Z., Albatal, R., Gurrin, C. G., & Smeaton, A. F. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-14442-9_29 · Cited by 4
Zhang, C., Wang, S., Liang, C., Liu, J., Huang, Q., Li, H., & Tian, Q. (2013) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/2502081.2502132 · Cited by 4
“Recently, the use of examplar image for object detection [16] and categorization [17] become popular.”
BackgroundA Real-Time Hand Pose Estimation System with Retrieval
Hou, G., Cui, R., & Zhang, C. (2015) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc.2015.305 · Cited by 4
“support vector machine method [17] with histogram of oriented gradients (HOG) [18] to detect the start pose.”
MethodUnsupervised Domain Adaptation Using Exemplar‐SVMs with Adaptation Regularization
He, Y., Tian, Y., Tang, J., & Ma, Y. (2018) · Complexity · doi:10.1155/2018/8425821 · Cited by 4
“In the opinion of [16], a discriminative exemplar classifier can be taken as a representation of a positive instance.”
BackgroundMethodInfluentialImage-Based 3D Model Retrieval for Indoor Scenes by Simulating Scene Context
Liu, M., Zhang, Y., He, J., Guo, J. J., & Guo, Y. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451547 · Cited by 4
“Exemplar-SVMs [12] are then used to match the synthesized scenes with the input image.”
Method
Metadata
- Authors
- Malisiewicz, T., Gupta, A., & Efros, A. A.
- Affiliations
- Carnegie Mellon University (US)
- Venue
- 2011 International Conference on Computer Vision
- Publisher
- IEEE
- Published
- 2011-11
- Pages
- 89-96
- OpenAlex
- W1989684337
- Cited by
- 865
- References
- 25
- Citations by year
- 2026: 4 · 2025: 18 · 2024: 16 · 2023: 29 · 2022: 54 · 2021: 52 · 2020: 49 · 2019: 45 · 2018: 65 · 2017: 78 · 2016: 105 · 2015: 148 · 2014: 97 · 2013: 65 · 2012: 36
- FWCI
- 48.23
- Type
- conference paper
- Language
- en
- Open access
- green
- figshare.com
- Licence
- cc-by
- Locations
- IEEE/CVF International Conference on Computer Vision (ICCV) (publishedVersion)
- Figshare (submittedVersion, cc-by)
- KiltHub Repository (submittedVersion)
- KiltHub Repository
- Funders
- National Science Foundation (0905402)
- Multidisciplinary University Research Initiative
- 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
- object detection, linear SVM classifier, Pascal VOC, scene understanding, part-based models, generalization
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.