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
Object Detection in 20 Years: A Survey
Zou, Z., Chen, K., Shi, Z., Guo, Y., & Ye, J. (2023) · Proceedings of the IEEE · doi:10.1109/jproc.2023.3238524 · Cited by 3,689
3D Object Representations for Fine-Grained Categorization
Krause, J., Stark, M., Deng, J., & Fei-Fei, L. (2013) · 2013 IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2013.77 · Cited by 3,496
Unsupervised Feature Learning via Non-parametric Instance Discrimination
Wu, Z., Xiong, Y., Yu, S. X., & Lin, D. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00393 · Cited by 3,664
Class-Balanced Loss Based on Effective Number of Samples
Cui, Y., Jia, M., Lin, T.-Y., Song, Y., & Belongie, S. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00949 · Cited by 3,094
“Another line of important work aims to study sample difficulty in terms of loss and assign higher weights to hard examples [11, 29, 8, 26].”
BackgroundTraining Region-Based Object Detectors with Online Hard Example Mining
Shrivastava, A., Gupta, A., & Girshick, R. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.89 · Cited by 2,686
“The idea of dataset boot-strapping [25, 30], typically called hard negative mining in recent work [12], appears in the training of most successful object detectors [7, 9, 12, 13, 15, 16, 23, 25, 29].”
BackgroundSelf-citationDeep Clustering for Unsupervised Learning of Visual Features
Caron, M., Bojanowski, P., Joulin, A., & Douze, M. (2018) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-01264-9_9 · Cited by 2,674
Food-101 – Mining Discriminative Components with Random Forests
Bossard, L., Guillaumin, M., & Van Gool, L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10599-4_29 · Cited by 2,072
“To obtain components, we could use these discriminative samples directly in spirit of exemplar svms [26].”
Method3D Semantic Parsing of Large-Scale Indoor Spaces
Armeni, I., Sener, O., Zamir, A. R., Jiang, H., Brilakis, I., Fischer, M., & Savarese, S. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.170 · Cited by 1,892
“3.2).”
InfluentialSUN RGB-D: A RGB-D scene understanding benchmark suite
Song, S., Lichtenberg, S. P., & Xiao, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298655 · Cited by 1,943
“For DPM and Exemplar SVM, we use the depth as another image channel and concatenate HOG computed from that and from color images.”
MethodInfluentialContrastive Learning for Unpaired Image-to-Image Translation
Park, T., Efros, A. A., Zhang, R., & Zhu, J.-Y. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58545-7_19 · Cited by 1,442
Self-citationScheirer, W. J., de Rezende Rocha, A., Sapkota, A., & Boult, T. E. (2013) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2012.256 · Cited by 1,573
Meta-Learning With Differentiable Convex Optimization
Lee, K., Maji, S., Ravichandran, A., & Soatto, S. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.01091 · Cited by 1,260
Domain Generalization: A Survey
Zhou, K., Liu, Z., Qiao, Y., Xiang, T., & Loy, C. C. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3195549 · Cited by 1,315
“As the ensemble of such exemplar SVMs have shown excellent generalization performance on the object detection task in [213], Xu et al.”
BackgroundLearning to Reweight Examples for Robust Deep Learning
Ren, M., Zeng, W., Yang, B., & Urtasun, R. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1803.09050 · Cited by 740
Cheng, G., Han, J., Zhou, P., & Guo, L. (2014) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2014.10.002 · Cited by 1,007
Vehicle detection in aerial imagery : A small target detection benchmark
Razakarivony, S., & Jurie, F. (2016) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2015.11.002 · Cited by 916
“…computed from integral histograms ( e.g. [60]), edgelets ( e.g. [51]), shapelets ( e.g. [61]), but also histogram of gradients (HOG) ( e.g. [4, 62]), bag-of-words ( e.g. [63]), multi-scale spatial pyramids [64], covariance descriptors ( e.g. [38]), co-occurrence features ( e.g. [65]), local…”
Dice Loss for Data-imbalanced NLP Tasks
Li, X., Sun, X., Meng, Y., Liang, J., Wu, F., & Li, J. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.45 · Cited by 646
Criminisi, A. (2011) · now publishers, Inc. eBooks · doi:10.1561/9781601985415 · Cited by 703
Meta-Weight-Net: Learning an Explicit Mapping For Sample Weighting
Shu, J., Xie, Q., Yi, L., Zhao, Q., Zhou, S., Xu, Z., & Meng, D. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.07379 · Cited by 385
“Hard example mining (Malisiewicz et al., 2011) and focal loss (Lin et al., 2018) also enhance the functions of harder examples and put larger weights on them.”
Advanced Deep-Learning Techniques for Salient and Category-Specific Object Detection: A Survey
Han, J., Zhang, D., Cheng, G., Liu, N., & Xu, D. (2018) · IEEE Signal Processing Magazine · doi:10.1109/msp.2017.2749125 · Cited by 621
“, locating the corresponding objects in the scene using known or unknown exemplars, which is the exemplar-based object detection [71]–[73]).”
BackgroundModeling video evolution for action recognition
Fernando, B., Gavves, E., Jose Oramas, M., Ghodrati, A., & Tuytelaars, T. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7299176 · Cited by 569
“Since we use parameters of a linear function as a new representation, our work also has some similarity to the exemplar SVM concept [18], but our objective is to learn a representation for the relative ordering of a set of frames in a video.”
MethodDu, G., Wang, K., Lian, S., & Zhao, K. (2021) · Artificial Intelligence Review · doi:10.1007/s10462-020-09888-5 · Cited by 475
“1 3 criterion Mirtich and Canny (1994). However, searching the grasp solution space is a complex problem which is quite time-consuming.”
BackgroundDeepEdge: A multi-scale bifurcated deep network for top-down contour detection
Bertasius, G., Shi, J., & Torresani, L. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7299067 · Cited by 525
“This hold-out dataset is used for the hard-positive mining step [22].”
MethodDeep Clustering for Unsupervised Learning of Visual Features
Caron, M., Bojanowski, P., Joulin, A., & Douze, M. (2018) · arXiv · doi:10.48550/arxiv.1807.05520 · Cited by 533
“Their approach discriminates between images in a similar way as examplar SVM [36], while we are simply clustering them.”
MethodExploration in deep reinforcement learning: A survey
Ladosz, P., Weng, L., Kim, M., & Oh, H. (2022) · Information Fusion · doi:10.1016/j.inffus.2022.03.003 · Cited by 525
“One example method is to use exemplar model [69] developed in [70].”
MethodWord Spotting and Recognition with Embedded Attributes
Almazan, J., Gordo, A., Fornes, A., & Valveny, E. (2014) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2014.2339814 · Cited by 517
Seeing 3D Chairs: Exemplar Part-Based 2D-3D Alignment Using a Large Dataset of CAD Models
Aubry, M., Maturana, D., Efros, A. A., Russell, B. C., & Sivic, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.487 · Cited by 473
“In Figure 5 we report full precision-recall curves for our algorithm and compare it against two baselines: (i) DPM [14] and (ii) a root template detector using the LDA version of Exemplar-SVM [28].”
BackgroundMethodInfluentialSelf-citationScatter Component Analysis: A Unified Framework for Domain Adaptation and Domain Generalization
Ghifary, M., Balduzzi, D., Kleijn, W. B., & Zhang, M. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2599532 · Cited by 471
“[22] extended an exemplar-SVM [61] to domain generalization by adding”
BackgroundLearning descriptors for object recognition and 3D pose estimation
Wohlhart, P., & Lepetit, V. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298930 · Cited by 358
“[19, 11, 28] show how to build discriminative models based on these representations using SVM or boosting applied to training data.”
BackgroundMethodActive Object Localization with Deep Reinforcement Learning
Caicedo, J. C., & Lazebnik, S. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.286 · Cited by 447
“A popular sliding window method, based on HOG templates and SVM classifiers, has been extensively used to localize objects [11, 21], parts of objects [8, 20], discriminative patches [29, 17] and even salient components of scenes [24].”
MethodMODEC: Multimodal Decomposable Models for Human Pose Estimation
Sapp, B., & Taskar, B. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.471 · Cited by 423
“Methods such as [11, 13] calibrate the models post-hoc using cross-validation data.”
BackgroundMethodLocal Aggregation for Unsupervised Learning of Visual Embeddings
Zhuang, C., Zhai, A., & Yamins, D. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00610 · Cited by 417
“[55, 49, 14, 75]) do not benefit significantly from increasing depth.”
BackgroundLearning feed-forward one-shot learners
Bertinetto, L., Henriques, J. F., Valmadre, J., Torr, P. H. S., & Vedaldi, A. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.05233 · Cited by 226
“These problems are usually tackled by using generative models [18, 13] or, in a discriminative setting, using ad-hoc solutions such as exemplar support vector machines (SVMs) [14].”
BackgroundBlocks That Shout: Distinctive Parts for Scene Classification
Juneja, M., Vedaldi, A., Jawahar, C. V., & Zisserman, A. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.124 · Cited by 368
“This approach is motivated by [17], that showed that a single example is often sufficient to train a reasonable, if a little restrictive, detector.”
BackgroundSliding Shapes for 3D Object Detection in Depth Images
Song, S., & Xiao, J. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10599-4_41 · Cited by 359
“Image-Based Detection: Popular detectors typically train a classifier on image area within a window, and test using the classifier via a sliding window [2, 3, 27] or on”
BackgroundMethodInfluentialLearning Deep Object Detectors from 3D Models
Peng, X., Sun, B., Ali, K., & Saenko, K. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.151 · Cited by 338
Kellenberger, B., Marcos, D., & Tuia, D. (2018) · Remote Sensing of Environment · doi:10.1016/j.rse.2018.06.028 · Cited by 335
“We propose to do so by means of hard negative mining (Malisiewicz et al., 2011; Shrivastava et al., 2016).”
MethodData-driven 3D Voxel Patterns for object category recognition
Xiang, Y., Choi, W., Lin, Y., & Savarese, S. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298800 · Cited by 319
“Nearest neighbor based methods [26] and deep neural networks [22, 18] handle the above factors in object category recognition implicitly.”
MethodRank Pooling for Action Recognition
Fernando, B., Gavves, E., Oramas M., J., Ghodrati, A., & Tuytelaars, T. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2558148 · Cited by 304
Subcategory-Aware Convolutional Neural Networks for Object Proposals and Detection
Xiang, Y., Choi, W., Lin, Y., & Savarese, S. (2017) · 2017 IEEE Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv.2017.108 · Cited by 298
Khanam, R., Hussain, M., Hill, R., & Allen, P. (2024) · IEEE Access · doi:10.1109/access.2024.3425166 · Cited by 237
“The HOG detector has become an integral component of numerous OD algorithms [172], [173], [174], and has found widespread applications in various CV tasks.”
MethodA Hierarchical Representation for Future Action Prediction
Lan, T., Chen, T.-C., & Savarese, S. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10578-9_45 · Cited by 292
“Instead of us-ing the high-dimensional HOG representation for clustering, we train an exemplar SVM [13,20] for each person example, and use the detection score of each example to create a K × K similarity matrix.”
MethodDeep Learning of Local RGB-D Patches for 3D Object Detection and 6D Pose Estimation
Kehl, W., Milletari, F., Tombari, F., Ilic, S., & Navab, N. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46487-9_13 · Cited by 283
“Other recent approaches [21,10,3] build discriminative models based on such representations using SVM or boosting applied to training data.”
BackgroundSafety Helmet Detection Based on YOLOv5
Zhou, F., Zhao, H., & Nie, Z. (2021) · 2021 IEEE International Conference on Power Electronics, Computer Applications (ICPECA) · doi:10.1109/icpeca51329.2021.9362711 · Cited by 272
“(3 ) Use classifiers for recognition, such as SVM [16] models.”
MethodNearest neighbors distance ratio open-set classifier
Mendes Júnior, P. R., de Souza, R. M., Werneck, R. D. O., Stein, B. V., Pazinato, D. V., de Almeida, W. R., Penatti, O. A. B., Torres, R. D. S., & Rocha, A. (2017) · Machine Learning · doi:10.1007/s10994-016-5610-8 · Cited by 276
“In many works in the literature, in spite of the explicit discrimination between classification and recognition, authors perform tests on recognition problems considering a closed-set scenario instead of an open-set one (Bartlett andWegkamp 2008; Chew et al. 2012; Jayadeva and Chandra 2007;Malisiewicz et al. 2011).”
BackgroundMethodHOGgles: Visualizing Object Detection Features
Vondrick, C., Khosla, A., Malisiewicz, T., & Torralba, A. (2013) · 2013 IEEE International Conference on Computer Vision · doi:10.1109/iccv.2013.8 · Cited by 282
“Consider the top detections for the exemplar object detector [9, 15] for a few images shown in Figure 5.”
BackgroundSelf-citationSemantic segmentation using regions and parts
Arbelaez, P., Hariharan, B., Gu, C., Gupta, S., Bourdev, L., & Malik, J. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248077 · Cited by 274
Discriminative Decorrelation for Clustering and Classification
Hariharan, B., Malik, J., & Ramanan, D. (2012) · Lecture Notes in Computer Science · doi:10.1007/978-3-642-33765-9_33 · Cited by 261
“This is in spite of the fact that our first-stage detectors require no training at all, and our second stage rescoring functions have an order of magnitude fewer parameters than ESVM+Co-occ [4] (for instance, for the horse category, in the second stage we have fewer than 2000 parameters, while ESVM+Co-occ has more tha…”
BackgroundMethodResultInfluentialAutomatically characterizing places with opportunistic crowdsensing using smartphones
Chon, Y., Lane, N. D., Li, F., Cha, H., & Zhao, F. (2012) · Proceedings of the 2012 ACM Conference on Ubiquitous Computing · doi:10.1145/2370216.2370288 · Cited by 270
“To recognize a variety of everyday objects observed within places, CSP adopts the examplarsvm approach proposed in [21].”
MethodJamal, M. A., Brown, M., Yang, M.-H., Wang, L., & Gong, B. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00763 · Cited by 247
Unsupervised Feature Learning via Non-Parametric Instance-level Discrimination
Wu, Z., Xiong, Y., Yu, S. X., & Lin, D. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.01978 · Cited by 163
“We take the class-wise supervision to the extreme of instance-wise supervision, and ask: Can we learn a meaningful metric that reflects apparent similarity among instances via pure discriminative learning? An image is distinctive in its own right, and each could differ significantly from other images in the same seman…”
BackgroundDo We Need More Training Data?
Zhu, X., Vondrick, C., Fowlkes, C. C., & Ramanan, D. (2016) · International Journal of Computer Vision · doi:10.1007/s11263-015-0812-2 · Cited by 204
“Specifically, they share an attractive property of exemplar SVMs (ESVMs) (Malisiewicz et al, 2011): each detection can be affiliated with its closest matching training example (given by the mixture index), allowing us to transfer annotations from a training example to the test instance.”
BackgroundResultInfluentialHistograms of Sparse Codes for Object Detection
Ren, X., & Ramanan, D. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.417 · Cited by 239
“Most approaches have explored model structure, either through non-parametric mixtures or exemplars [21, 10], compositional grammar structure [15], supervised correspondences [5, 2], and low-dimensional projections [29, 25].”
BackgroundMethodWeb-scale training for face identification
Taigman, Y., Yang, M., Ranzato, M., & Wolf, L. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298891 · Cited by 227
“In [16] multiple SVM classifiers, each based on a single positive sample, are used to construct a powerful descriptor of the learned class.”
MethodBabenko, A., & Lempitsky, V. (2012) · 2012 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2012.6248038 · Cited by 200
“Furthermore, it is straightforward to replace the (square of the) Euclidean distance within the multi-index with any other additive distance measure or kernel; it will thus be interesting to evaluate inverted multiindices within large-scale machine learning systems, in particular those utilizing exemplar SVMs [41].”
BackgroundAudio-Visual Instance Discrimination with Cross-Modal Agreement
Morgado, P., Vasconcelos, N., & Misra, I. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.01229 · Cited by 211
“The learned representations are optimized for ‘instance discrimination’ [17, 51, 86], i.”
BackgroundMethodSharma, V. K., & Mir, R. N. (2020) · Computer Science Review · doi:10.1016/j.cosrev.2020.100301 · Cited by 221
Wu, X., Hong, D., Tian, J., Chanussot, J., Li, W., & Tao, R. (2019) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2019.2897139 · Cited by 212
“2) Overview of Performance Comparison: To quantitatively assess the detection performances of the proposed method, we compare several state-of-the-art methods related to our framework, such as exemplar-SVMs [56], rotation-aware features [57], collection of part detectors-based [24], bag-ofwords with SVM (BOW-SVM) [58]…”
MethodData-driven visual similarity for cross-domain image matching
Shrivastava, A., Malisiewicz, T., Gupta, A., & Efros, A. A. (2011) · Proceedings of the 2011 SIGGRAPH Asia Conference · doi:10.1145/2024156.2024188 · Cited by 209
“…recent work in visual recognition has shown that it’s possible to train a discriminative classifier using a single positive instance and a large body of negatives [Wolf et al. 2009; Malisiewicz et al. 2011], provided that the negatives do not contain any images similar to the positive instance.”
BackgroundMethodInfluentialSelf-citationContextual Hypergraph Modeling for Salient Object Detection
Li, X., Li, Y., Shen, C., Dick, A., & Hengel, A. V. D. (2013) · 2013 IEEE International Conference on Computer Vision · doi:10.1109/iccv.2013.413 · Cited by 205
“We can use other classifiers such as the exemplar SVM [25], where the standard hinge-loss SVM is used.”
MethodInfluence-Balanced Loss for Imbalanced Visual Classification
Park, S., Lim, J., Jeon, Y., & Choi, J. Y. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00077 · Cited by 192
“Recently, numerous studies have been conducted in which each sample is considered to design sample-wise loss functions [9, 22, 27].”
BackgroundMethodFinding Things: Image Parsing with Regions and Per-Exemplar Detectors
Tighe, J., & Lazebnik, S. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.386 · Cited by 207
“Following the per-exemplar framework of [19], we train a per-exemplar detector for each labeled object instance in our dataset.”
BackgroundMethodInfluentialBabenko, A., & Lempitsky, V. (2015) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2014.2361319 · Cited by 190
Bertasius, G., Shi, J., & Torresani, L. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.65 · Cited by 189
“We use this for the hard-positive mining [21] in order to reduce the number of false-negative predictions.”
MethodDeformable Part Descriptors for Fine-Grained Recognition and Attribute Prediction
Zhang, N., Farrell, R., Iandola, F., & Darrell, T. (2013) · 2013 IEEE International Conference on Computer Vision · doi:10.1109/iccv.2013.96 · Cited by 193
“Following recent work by Belhumeur et al. [5] on localizing fiducial points in faces, Liu et al. [29] present an alternate approach to build an exemplar-based geometric and appearance models for detecting dog faces and localizing the facial keypoints.”
BackgroundSuperparsing: Scalable Nonparametric Image Parsing with Superpixels
Tighe, J., & Lazebnik, S. (2013) · International Journal of Computer Vision · doi:10.1007/s11263-012-0574-z · Cited by 193
“Most of these methods operate with a few predefined classes and require a generative or discriminative model to be trained in advance for each class (and sometimes even for each training exemplar [28, 29]).”
BackgroundBSL-1K: Scaling Up Co-articulated Sign Language Recognition Using Mouthing Cues
Albanie, S., Varol, G., Momeni, L., Afouras, T., Chung, J. S., Fox, N., & Zisserman, A. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58621-8_3 · Cited by 176
Exploiting Low-Rank Structure from Latent Domains for Domain Generalization
Xu, Z., Li, W., Niu, L., & Xu, D. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10578-9_41 · Cited by 190
“In contrast to [25], we introduce a nuclear norm based regularizer on the likelihood matrix in order to exploit the low-rank structure from latent domains for domain generalization.”
MethodResultAutomatic Ship Classification from Optical Aerial Images with Convolutional Neural Networks
Gallego, A.-J., Pertusa, A., & Gil, P. (2018) · Remote Sensing · doi:10.3390/rs10040511 · Cited by 168
“[60] based on a set of exemplar-based SVMs.”
Method2D-Driven 3D Object Detection in RGB-D Images
Lahoud, J., & Ghanem, B. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.495 · Cited by 177
“Multiple techniques have attempted to extend 2D detections into 3D using a single image [12, 13, 21], but these techniques need prior knowledge of the scene and do not genFigure 1.”
BackgroundClothing Co-parsing by Joint Image Segmentation and Labeling
Yang, W., Luo, P., & Lin, L. (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.2014.407 · Cited by 179
“The first phase, referred as “image co-segmentation”, iterates to extract consistent regions on images and jointly refines the regions over all images by employing the exemplar-SVM (ESVM) technique [23].”
BackgroundMethodInfluentialFew-Example Object Detection with Model Communication
Dong, X., Zheng, L., Ma, F., Yang, Y., & Meng, D. (2019) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2018.2844853 · Cited by 167
“The algorithm of [53] is based on the linear SVM classifier, which is limited to using the offthe-shelf features.”
MethodDiscriminatively Trained Templates for 3D Object Detection: A Real Time Scalable Approach
Rios-Cabrera, R., & Tuytelaars, T. (2013) · 2013 IEEE International Conference on Computer Vision · doi:10.1109/iccv.2013.256 · Cited by 169
“An example of such methods, trained to handle templates and metadata transfer, was proposed recently by Malisiewicz [9].”
MethodHardness-Aware Deep Metric Learning
Zheng, W., Chen, Z., Lu, J., & Zhou, J. (2019) · 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.2019.00016 · Cited by 189
“Hard Negative Mining: Hard negative mining has been employed in many machine learning tasks to enhance the training efficiency and boost performance, like supervised learning [25, 13, 46, 10, 45], exemplar based learning [21] and unsupervised learning [40, 1].”
BackgroundKraft, M., Piechocki, M., Ptak, B., & Walas, K. (2021) · Remote Sensing · doi:10.3390/rs13050965 · Cited by 187
Lu, E., Xie, W., & Zisserman, A. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-20893-6_42 · Cited by 210
Clothes Co-Parsing Via Joint Image Segmentation and Labeling With Application to Clothing Retrieval
Liang, X., Lin, L., Yang, W., Luo, P., Huang, J., & Yan, S. (2016) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2016.2542983 · Cited by 166
“where NE indicates the negative examples, and f(·) is the appearance feature of a region, following [1].”
BackgroundMethodInfluentialFrom Virtual to Reality: Fast Adaptation of Virtual Object Detectors to Real Domains
Sun, B., & Saenko, K. (2014) · Proceedings of the British Machine Vision Conference 2014 · doi:10.5244/c.28.82 · Cited by 157
Piecewise Classifier Mappings: Learning Fine-Grained Learners for Novel Categories With Few Examples
Wei, X.-S., Wang, P., Liu, L., Shen, C., & Wu, J. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2924811 · Cited by 159
Template Adaptation for Face Verification and Identification
Crosswhite, N., Byrne, J., Stauffer, C., Parkhi, O., Cao, Q., & Zisserman, A. (2017) · Image and Vision Computing · doi:10.1016/j.imavis.2018.09.002 · Cited by 173
McAllister, P. M., Zheng, H., Bond, R., & Moorhead, A. (2018) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2018.02.008 · Cited by 166
Data-driven visual similarity for cross-domain image matching
Shrivastava, A., Malisiewicz, T., Gupta, A. K., & Efros, A. A. (2011) · ACM Transactions on Graphics · doi:10.1145/2070781.2024188 · Cited by 172
Self-citationDetecting animals in African Savanna with UAVs and the crowds
Rey, N., Volpi, M., Joost, S., & Tuia, D. (2017) · Remote Sensing of Environment · doi:10.1016/j.rse.2017.08.026 · Cited by 150
“Despite these challenges, recent developments in object detection pipelines in both computer vision (Malisiewicz et al., 2011; Girshick et al., 2014) and remote sensing (Tuermer et al., 2013; Moranduzzo and Melgani, 2014; Akc¸ay and Aksoy, 2016), provide promising techniques to semi-automatically…”
BackgroundMethodInfluentialContext Driven Scene Parsing with Attention to Rare Classes
Yang, J., Price, B., Cohen, S., & Yang, M.-H. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.415 · Cited by 154
“With the same concern on interesting object classes, Tighe and Lazebnik [21] propose to augment their previous superpixel based parsing system [22] with pre-trained exemplar SVMs [16].”
BackgroundExemplar-Based Color Constancy and Multiple Illumination
Joze, H. R. V., & Drew, M. S. (2013) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2013.169 · Cited by 156
Clothing Co-Parsing by Joint Image Segmentation and Labeling
Yang, W. P., Luo, P., & Lin, L. (2014) · Cited by 147
Underwater object detection using Invert Multi-Class Adaboost with deep learning
Chen, L., Liu, Z., Tong, L., Jiang, Z., Wang, S., Dong, J., & Zhou, H. (2020) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn48605.2020.9207506 · Cited by 159
“For example, focal loss [16] and hard example mining [17] emphasise samples with higher training loss while self-placed learning [18] and curriculum learning [19] encourage learning samples with low loss.”
BackgroundCapturing Long-Tail Distributions of Object Subcategories
Zhu, X., Anguelov, D., & Ramanan, D. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.122 · Cited by 132
“Our long-tail models with sharing considerably outperform such methods in Table 1, 0.8 0.85 0.9 0.950.8 0.85 0.9 0.95 1 Cat head detection Our 6−DPM Shape+texture [24][ 5] Recall Pr ec is io n 0.2 0.4 0.6 0.8 10.2 0.4 0.6 0.8 1 Dog face detection Our 6−DPM SVM+RBF+SIFT [14][15] Recall Figure 7: We run our implementati…”
BackgroundMethodInfluentialImageNet Auto-Annotation with Segmentation Propagation
Guillaumin, M., Kuettel, D., & Ferrari, V. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-014-0713-9 · Cited by 173
“Malisiewicz et al. [44] proposes to employ per-exemplar SVMs to find neighbours for transfer, instead of simply measuring appearance similarity.”
BackgroundRepresenting Videos Using Mid-level Discriminative Patches
Jain, A., Gupta, A. K., Rodriguez, M. D., & Davis, L. S. (2013) · 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.2013.332 · Cited by 137
“Finally, we re-rank the clusters using the approach described next.”
BackgroundMethodSelf-citationWu, X., Hong, D., Chanussot, J., Xu, Y., Tao, R., & Wang, Y. (2019) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2019.2919755 · Cited by 137
“To effectively evaluate the performances of the proposed method (FRIFB), we make a comparison with some state-of-the-art algorithms: bag-of-words (BOW)-SVM [13], rotation-aware features [14], exemplar-SVMs [15], the collection of part detectors (COPD) [12], rotation dense feature pyramid networks [10], rotation-invari…”
MethodScene Parsing with Object Instances and Occlusion Ordering
Tighe, J., Niethammer, M., & Lazebnik, S. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.479 · Cited by 133
“Then it computes two pixel-level potentials: a region-based data term, based on a nonparametric voting score for similar regions in the retrieval set; and a detector-based data term, obtained from responses of per-exemplar detectors [16] corresponding to instances in the retrieval set.”
MethodImproving Generalization via Scalable Neighborhood Component Analysis
Wu, Z., Efros, A. A., & Yu, S. X. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01234-2_42 · Cited by 163
Self-citationUnderstanding the coverage and scalability of place-centric crowdsensing
Chon, Y., Lane, N. D., Kim, Y., Zhao, F., & Cha, H. (2013) · ACM International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) · doi:10.1145/2493432.2493498 · Cited by 135
“Object recognition exemplar-SVM [23] bicycle, sofa, bottle, bus, car, chair, dining table, person, potted plant, monitor PASCAL VOC 2011 dataset [14]”
Learning from Demonstrations Through the Use of Non-rigid Registration
Schulman, J., Ho, J., Lee, C., & Abbeel, P. (2016) · Springer Tracts in Advanced Robotics · doi:10.1007/978-3-319-28872-7_20 · Cited by 133
“Also in vision, using nearest-neighbor based techniques (not involving registration) to transfer various sorts of metadata from one image onto another has had some recent success [4–6].”
MethodIndoor Scene Understanding in 2.5/3D for Autonomous Agents: A Survey
Naseer, M., Khan, S. S., & Porikli, F. (2018) · IEEE Access · doi:10.1109/access.2018.2886133 · Cited by 156
“Both processed images are then fed to the AlexNet to extract features which are then concatenated before passing through a SVM classifier for object detection and a support vector regressor (SVR) for pose estimation.”
MethodInfluentialA critical analysis of self-supervision, or what we can learn from a single image
Asano, Y. M., Rupprecht, C., & Vedaldi, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.13132 · Cited by 95
“Single sample learning was pursued by the semi-parametric exemplar SVM model [34].”
MethodNature Conservation Drones for Automatic Localization and Counting of Animals
van Gemert, J. C., Verschoor, C., Mettes, P., Epema, K., Koh, L. P., & Wich, S. A. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-16178-5_17 · Cited by 138
Graph-Based Discriminative Learning for Location Recognition
Cao, S., & Snavely, N. (2013) · 2013 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2013.96 · Cited by 129
Painting-to-3D model alignment via discriminative visual elements
Aubry, M., Russell, B. C., & Šivic, J. (2014) · ACM Transactions on Graphics · doi:10.1145/2591009 · Cited by 120
“A visual element detector is defined as a linear classifier trained from a single patch as a positive example and a large number of “background” patches as negative examples, similar to an exemplar Support Vector Machine (SVM) [Malisiewicz et al. 2011; Shrivastava et al. 2011].”
Influential
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.