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 401–500 of 873
Bolstering Maritime Object Detection with Synthetic Data
Becktor, J., Schöller, F. E. T., Boukas, E., Blanke, M., & Nalpantidis, L. (2022) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2022.10.410 · Cited by 15
ASIST: Automatic semantically invariant scene transformation
Litany, O., Remez, T., Freedman, D., Shapira, L., Bronstein, A. M., & Gal, R. (2016) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2016.08.002 · Cited by 20
“me requirements and the strength of the GPU. Song et al. [6] render a large dataset of models from multiple angles and train a linear SVM classifier for each, according to the exemplar SVM approach of [9]. Using a sliding window over a scene, they run each of the classifiers and select a small number of possible models,…”
MethodMulti-class Semantic Video Segmentation with Exemplar-Based Object Reasoning
Liu, B., He, X., & Gould, S. J. (2015) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2015.140 · Cited by 14
Vector array based Multi-View Face Detection with compound exemplars
Ma, K., & Ben-Arie, J. (2012) · 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.2012.6248053 · Cited by 14
“Based on a much smaller training set, our system achieves better results than the current state-ofthe-art object detection systems [9, 19] with a reasonable”
BackgroundMethodResultInfluentialMultiple-Kernel, Multiple-Instance Similarity Features for Efficient Visual Object Detection
Sun, C., & Lam, K. (2013) · IEEE Transactions on Image Processing · doi:10.1109/tip.2013.2255303 · Cited by 14
Boosted Random Ferns for Object Detection
Villamizar, M., Andrade‐Cetto, J., Sanfeliu, A., & Moreno-Noguer, F. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2676778 · Cited by 17
“SVM [2], [7], [10], [20], [24], [34] and Boosting [25], [36], [44], [49], [58], [61], [64] are typical classifier used for this purpose.”
BackgroundKernel Square-Loss Exemplar Machines for Image Retrieval
de Rezende, R. S., Zepeda, J., Ponce, J., Bach, F., & Pérez, P. (2017) · 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.2017.768 · Cited by 14
“[24], leverages the availability of large, unannotated pools of images within the context of supervised learning.”
BackgroundMethodInfluentialUnderwater Fish Identification in Real-Time using Convolutional Neural Network
K, A. S., Deepa, N., & Devi., T. (2023) · International Conference on Intelligent Computing and Control Systems (ICICCS) · doi:10.1109/iciccs56967.2023.10142531 · Cited by 14
“3: starting point the item weights Ij in order Use to calculate the weights of samples tested (15)”
BackgroundBuilding a dictionary of image fragments
Liao, Z., Farhadi, A., Wang, Y., Endres, I., & Forsyth, D. A. (2012) · 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.2012.6248085 · Cited by 14
“There exist methods for image classification using region-based exemplar matching [5, 7] and image-based exemplar matching [4, 12].”
MethodResultInfluentialInteractive adaptation of real-time object detectors
Goehring, D., Hoffman, J., Rodner, E., Saenko, K., & Darrell, T. (2014) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra.2014.6907018 · Cited by 14
Adaptive pedestrian detection by predicting classifier
Tang, S. Y., Ye, M., Xu, P., & Li, X. (2017) · Neural Computing and Applications · doi:10.1007/s00521-017-3152-z · Cited by 14
“Here the proprietary classifier uses Exemplar SVM (ESVM) which is very effective in object detection [18].”
MethodTan, W., Yan, B., & Lin, C. (2017) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2017.2742243 · Cited by 14
“We choose linear SVM to learn weights of visual words (actually learning features’ weight because the weights of visual words will finally be mapped back to features, see the following subsection) for two reasons: (i) a linear SVM generalizes better even with a single positive sample and a very large negative set of s…”
BackgroundMethodLearning Local Invariant Mahalanobis DistancesPreprint
Fetaya, E., & Ullman, S. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1502.01176 · Cited by 14
“The exemplar-SVM algorithm [18] can be seen as a local similarity measure.”
MethodThree viewpoints toward exemplar SVM
Kobayashi, T. (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.2015.7298893 · Cited by 15
“reconstruction (original definition [17])”
BackgroundMethodInfluentialDeep Boosting Learning: A Brand-New Cooperative Approach for Image-Text Matching
Diao, H., Zhang, Y., Gao, S., Ruan, X., & Lu, H. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3396063 · Cited by 13
“This is inspired by the analysis of minimizing a modified non-trivial loss function with uniform sampling in classification tasks [67], [68], identification tasks [54], [55], and multi-modal tasks [18], [69].”
MethodQiu, S., Wen, G., Deng, Z., Fan, Y., & Hui, B. (2017) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2017.2731863 · Cited by 14
“We take PCM [1], MS-CNN [18], DPM [11], and ESVM [19] as our baselines.”
MethodFedCSS: Joint Client-and-Sample Selection for Hard Sample-Aware Noise-Robust Federated Learning
Li, A., Cao, Y., Guo, J., Peng, H., Guo, Q., & Yu, H. (2023) · Proceedings of the ACM on Management of Data · doi:10.1145/3617332 · Cited by 14
Using spatiotemporal information in weather radar data to detect and track communal roosts
Pérez, G., Zhao, W., Cheng, Z., Belotti, M. C. T. D., Deng, Y., Simons, V. F., Tielens, E. K., Kelly, J. F., Horton, K. G., Maji, S., et al. (2024) · Remote Sensing in Ecology and Conservation · doi:10.1002/rse2.388 · Cited by 13
Recognition Using Hybrid Classifiers
Osadchy, M., Keren, D., & Raviv, D. (2015) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2015.2465910 · Cited by 16
Self-Training with Differentiable Teacher
Zuo, S., Yu, Y., Chen, L., Jiang, H., Er, S., Zhang, C., Zhao, T., & Hongyuan Zha, B. (2022) · Findings of the Association for Computational Linguistics: NAACL · doi:10.18653/v1/2022.findings-naacl.70 · Cited by 14
Large-scale interactive retrieval in art collections using multi-style feature aggregation
Ufer, N., Simon, M., Lang, S., & Ommer, B. (2021) · PLoS ONE · doi:10.1371/journal.pone.0259718 · Cited by 15
Deep Attributes for One-Shot Face Recognition
Jadhav, A., Namboodiri, V. P., & Venkatesh, K. S. (2016) · Lecture notes in computer science · doi:10.1007/978-3-319-49409-8_44 · Cited by 13
“This was observed for two different methods, one-shot recognition using Exemplar-SVM based and oneshot similarity kernel based techniques.”
MethodInfluentialExemplar-SVMs for visual object detection, label transfer and image retrieval
Malisiewicz, T., Shrivastava, A., Gupta, A., & Efros, A. A. (2012) · Cited by 13
Self-citationReweighting Augmented Samples by Minimizing the Maximal Expected LossPreprint
Yi, M., Hou, L., Shang, L., Jiang, X., Liu, Q., & Ma, Z.-M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.08933 · Cited by 23
“Similarly, hard example mining (Malisiewicz et al., 2011) downsamples the majority class and exploits the most difficult examples.”
BackgroundEfficient Oil Tank Detection Using Deep Learning: A Novel Dataset and Deployment on Edge Devices
Rizk, M., & Chehade, A. (2024) · IEEE Access · doi:10.1109/access.2024.3495523 · Cited by 12
“The adoption of machine learning techniques, notably Support Vector Machines (SVM) [36], has signified a transformative shift towards more sophisticated classification and detection capabilities.”
Multi-Max-Margin Support Vector Machine for multi-source human action recognition
Wu, D., & Shao, L. (2013) · Neurocomputing · doi:10.1016/j.neucom.2013.08.038 · Cited by 12
Knowing a Good HOG Filter When You See It: Efficient Selection of Filters for Detection
Ahmed, E., Shakhnarovich, G., & Maji, S. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-10590-1_6 · Cited by 12
Improving the search accuracy of the VLAD through weighted aggregation of local descriptors
Kim, T.-E., & Kim, M. H. (2015) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2015.07.005 · Cited by 12
Exploiting Object Similarity in 3D Reconstruction
Zhou, C., Güney, F., Wang, Y., & Geiger, A. (2015) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.254 · Cited by 12
“, building or car), we train an ensemble of linear exemplar SVMs [29] directly on the TSDF volume using a small (≤ 3) number of annotated instances.”
MethodInfluentialEXMOVES: Mid-level Features for Efficient Action Recognition and Video Analysis
Tran, D., & Torresani, L. (2016) · International Journal of Computer Vision · doi:10.1007/s11263-016-0905-6 · Cited by 12
“Adopting the exemplar-SVM method in Malisiewicz et al. (2011), our exemplar-SVM training procedure learns a linear classifier f (x) = wTφ(x) + b, by minimizing the following objective function: min w,b ‖w‖2 + C1 ∑ x∈V+ s.t. |x∩xE ||xE | ≥0.5 h ( wTφ(x) + b) +C2 N∑ i=1 ∑ x∈V−i h (−wTφ(x) − b)…”
BackgroundMethodInfluentialEffective 3D object detection based on detector and tracker
Nie, W., Liu, A.-A., Wang, Z., & Su, Y. (2016) · Neurocomputing · doi:10.1016/j.neucom.2015.06.118 · Cited by 12
Learning to Reweight with Deep Interactions
Yang, F., Xia, Y., Wu, L., Xie, S., Liu, W., Bian, J., Qin, T., & Li, X. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i8.16906 · Cited by 15
“…al. 2000) where the weights of incorrectly classi-fied data are gradually increased, or implicitly achieved by controlling the sampling probability, like hard negative mining (Malisiewicz, Gupta, and Efros 2011) where the harder examples in a previous round will be sampled again in the next round.”
BackgroundClick-Based Student Performance Prediction: A Clustering Guided Meta-Learning Approach
Chu, Y.-W., Tenorio, E., Castro, L. M. C., Douglas, K. A., Lan, A., & Brinton, C. G. (2021) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata52589.2021.9671729 · Cited by 15
“A significant amount of research in meta-learning has focused on solving label imbalances through sample reweighting approaches [48]–[50].”
BackgroundBoosting masked dominant orientation templates for efficient object detection
Rìos-Cabrera, R., & Tuytelaars, T. (2014) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2013.12.008 · Cited by 12
Adapting a Pedestrian Detector by Boosting LDA Exemplar Classifiers
Xu, J., Vázquez, D., Ramos, S., López, A. M., & Ponsa, D. (2013) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2013.104 · Cited by 13
““TAR-FULL-SVM” represents the original HOG-SVM detector [3] which is trained with INRIA full training dataset.”
BackgroundMethodInfluentialScene text recognition using sparse coding based features
Zhang, D., Wang, D., & Wang, H. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025212 · Cited by 12
Joint image representation and classification in random semantic spaces
Zhang, C., Zhu, X., Li, L., Zhang, Y., Liu, J., Huang, Q., & Tian, Q. (2015) · Neurocomputing · doi:10.1016/j.neucom.2014.12.083 · Cited by 12
Multiple Human Identification and Cosegmentation: A Human-Oriented CRF Approach With Poselets
Zhu, H., Lu, J., Cai, J., Zheng, J., Lu, S., & Magnenat‐Thalmann, N. (2016) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2016.2571629 · Cited by 12
“Recently, Tighe and Lazebnik [44] utilized the rigid shape information transferred by Exemplar SVM [47] in scene understanding tasks, which achieved state-of-the-art performance.”
BackgroundGenerating Hard Examples for Pixel-Wise Classification
Lee, H., Kwon, H., & Kim, W. (2021) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2021.3112924 · Cited by 15
“Thereafter, hard example mining has been widely used in various applications such as pedestrian detection [17], [18], human pose estimation [19], [20], action recognition [21], [22], event recognition [23], object detection [24]–[27], and so on.”
MethodWhat Is Usual in Unusual Videos? Trajectory Snippet Histograms for Discovering Unusualness
İşcen, A., Armagan, A., & Duygulu, P. (2014) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2014.123 · Cited by 12
“Remaining snippets are used to construct initial models, and an examplar-SVM [12] is trained for each model.”
BackgroundMethodEstimation of Sunlight Direction Using 3D Object Models
Liu, Y., Gevers, T., & Li, X. (2014) · IEEE Transactions on Image Processing · doi:10.1109/tip.2014.2378032 · Cited by 12
“To achieve this, we combine the following contributions: (1) To locate objects using the DPM detector [9] and classify their category using exemplar-SVMs [10]; (2) To train exemplar-SVMs using (synthetic) images generated from 3D models and a wide range of viewpoints; (3) To generate”
BackgroundMethodInfluentialSemantic mapping and semantics-boosted navigation with path creation on a mobile robot
Sun, H., Meng, Z., & Ang, M. H. (2017) · IEEE Conference on Cybernetics and Intelligent Systems · doi:10.1109/iccis.2017.8274775 · Cited by 12
“Same as scene classification, hand-crafted features are heavily used in early stage [8], [9].”
Learning a Robust Society of Tracking Parts Using Co-occurrence Constraints
Burceanu, E., & Leordeanu, M. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-11009-3_9 · Cited by 12
Discriminative Video Representation Learning Using Support Vector Classifiers
Wang, J., & Cherian, A. (2019) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2019.2937292 · Cited by 13
“Similar works are Exemplar-SVMs [57], [58],”
Clustering-based multiple instance learning with multi-view feature
He, C., Shao, J., Zhang, J., & Zhou, X. (2019) · Expert Systems with Applications · doi:10.1016/j.eswa.2019.113027 · Cited by 13
Hara, K., Sun, J. Y., Chazan, J., Jacobs, D. W., & Froehlich, J. E. (2013) · Proceedings of the AAAI Conference on Human Computation and Crowdsourcing · doi:10.1609/hcomp.v1i1.13109 · Cited by 12
Discriminative Weighted Sparse Partial Least Squares for Human Detection
Li, Q., Yan, Y., & Wang, H. (2015) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2015.2495342 · Cited by 11
3D shape retrieval using a single depth image from low-cost sensors
Feng, J., Wang, Y., & Chang, S. (2016) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2016.7477652 · Cited by 11
“inspired from successful ideas using an ensemble of weak classifiers in detection and recognition [18][29][25].”
MethodAn Aircraft Object Detection Algorithm Based on Small Samples in Optical Remote Sensing Image
Wang, T., Cao, C. C., Zeng, X., Feng, Z., Shen, J., Li, W., Wang, B., Zhou, Y., & Xu, Y. H. (2020) · Applied Sciences · doi:10.3390/app10175778 · Cited by 11
“adopt the integration idea in machine learning, and train an SVM classifier for each positive sample by using HOG features [19].”
MethodSelf-supervised Secondary Landmark Detection via 3D Representation Learning
Bala, P., Zimmermann, J., Park, H. S., & Hayden, B. Y. (2023) · International Journal of Computer Vision · doi:10.1007/s11263-023-01804-y · Cited by 12
“Associated signals can be represented as an image with itself (Malisiewicz et al., 2011; Shrivastava et al., 2011; Dosovitskiy et al., 2016; He et al., 2020a; Wu et al., 2018), neighboring patches within an image (Isola et al., 2015; Hénaff et al., 2020; Oord & Li, 2018), or multiple views of the…”
BackgroundBlobCUT: A Contrastive Learning Method to Support Small Blob Detection in Medical Imaging
Li, T., Xu, Y., Wu, T., Charlton, J. R., Bennett, K. M., & Al-Hindawi, F. (2023) · Bioengineering · doi:10.3390/bioengineering10121372 · Cited by 11
“Associated signals can be an image with itself [38,41,44–46], an image with its downstream representation [47], neighboring patches within an image [35,39,48], or multiple views of the input image [49], and most successfully, an image with a set of transformed versions of itself [37,40].”
BackgroundLooking Inside Category: Subcategory-Aware Object Recognition
Dong, J., Chen, Q., Feng, J., Jia, K., Huang, Z., & Yan, S. (2014) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2014.2355697 · Cited by 12
“plane bike bird boat bottle bus car cat chair cow table dog horse motor person plant sheep sofa train tv mAP E-SVM [6] 20.”
BackgroundMethodInfluentialImmediate, Scalable Object Category Detection
Aytar, Y., & Zisserman, A. (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.2014.305 · Cited by 10
“Finally, we use the fast search for templates obtained from E-SVM models for VOC07, and compare the performance to the original E-SVM template and to Video Google (where for VG the query is the image BB that is used to determine the E-SVM).”
MethodInfluentialFuncISH: learning a functional representation of neural ISH images
Liscovitch, N., Shalit, U., & Chechik, G. (2013) · Bioinformatics · doi:10.1093/bioinformatics/btt207 · Cited by 11
Dynamic Selection of Exemplar-SVMs for Watch-list Screening through Domain Adaptation
Bashbaghi, S., Granger, É., Sabourin, R., & Bilodeau, G.-A. (2017) · International Conference on Pattern Recognition Applications and Methods · doi:10.5220/0006256507380745 · Cited by 11
“As a specialized classification technique considering the SSPP problem, e-SVM classifier is adapted using non-target video ROIs (Bashbaghi et al., 2016), (Malisiewicz et al., 2011).”
MethodLDA classifier monitoring in distributed streaming systems
Bernstein, R., Osadchy, M., Keren, D., & Schuster, A. (2018) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2018.09.017 · Cited by 11
Learning and Calibrating Per-Location Classifiers for Visual Place Recognition
Gronát, P., Šivic, J., Obozinski, G., & Pajdla, T. (2016) · International Journal of Computer Vision · doi:10.1007/s11263-015-0878-x · Cited by 13
“We develop a non-parametric procedure to calibrate the outputs of the large number of per-location classifiers without the need for additional positive training data.”
BackgroundMethodInfluentialContextual Hypergraph Modelling for Salient Object DetectionPreprint
Li, X., Li, Y., Shen, C., Dick, A. R., & van den Hengel, A. (2013) · arXiv (Cornell University) · doi:10.48550/arxiv.1310.5767 · Cited by 12
“We can use other classifiers such as the exemplar SVM [25], where the standard hinge-loss SVM is used.”
MethodAn end-to-end convolutional selective autoencoder approach to Soybean Cyst Nematode eggs detectionPreprint
Akintayo, A., Lee, N., Chawla, V., Mullaney, M. P., Marett, C. C., Singh, A. K., Singh, A., Tylka, G. L., Ganapathysubramaniam, B., & Sarkar, S. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.07834 · Cited by 11
“Therefore, a thresholded differencing was added to the non-maximum suppression scheme (Malisiewicz et al., 2011).”
BackgroundMethodInfluentialHow important are Deformable Parts in the Deformable Parts Model?Preprint
Divvala, S., Efros, A. A., & Hebert, M. (2012) · arXiv (Cornell University) · doi:10.1184/r1/6555053 · Cited by 11
Self-citationLearning to Teach with Deep Interactions.Preprint
Yang, F., Xia, Y., Wu, L., Xie, S., Liu, W., Bian, J., Qin, T., Li, X., & Liu, T. (2020) · arXiv (Cornell University) · Cited by 11
Low resolution remote sensing object detection with fine grained enhancement and swin transformer
Xu, Z., Wang, X., Huang, K., & Chen, R. (2025) · Scientific Reports · doi:10.1038/s41598-025-10286-6 · Cited by 12
Attribute-based vehicle recognition using viewpoint-aware multiple instance SVMs
Duan, K., Marchesotti, L., & Crandall, D. (2014) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2014.6836081 · Cited by 10
“They use exemplar SVMs [18] to find salient regions, while using click pair information to jointly infer object parts.”
MethodMulti-Resolution Language Grounding with Weak Supervision
Koncel-Kedziorski, R., Hajishirzi, H., & Farhadi, A. (2014) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.3115/v1/d14-1043 · Cited by 13
“An exemplar SVM is an SVM fit with one positive and many negative instances, allowing us to define an example by what it is not (Malisiewicz et al., 2011; Shrivastava et al., 2011).”
BackgroundObject Recognition for Manipulation Tasks in Real Domestic Settings: A Comparative Study
Martínez, L., Loncomilla, P., & Ruíz-del-Solar, J. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-18615-3_17 · Cited by 10
“achieve, and it requires recognizing the object’s boundaries [16], although global approaches like ensemble of exemplar HoG SVMs detectors [17] that use one positive example against multiple negative ones are an interesting alternative for solving this problem.”
BackgroundContext Forest for Object Class Detection
Modolo, D., Vezhnevets, A. S., & Ferrari, V. (2015) · British Machine Vision Conference (BMVC) · doi:10.5244/c.29.188 · Cited by 10
“These detectors [4, 14, 15, 17, 21, 26, 36, 60] model an object class as mixture of components.”
BackgroundMethodInfluentialAction recognition by discriminative EdgeBoxes
El‐Masry, M., Fakhr, M. W., & Salem, M. A. ‐. (2017) · IET Computer Vision · doi:10.1049/iet-cvi.2017.0335 · Cited by 10
Towards a self-sufficient face verification system
López‐López, E., Regueiro, C. V., Pardo, X. M., Franco, A., & Lumini, A. (2021) · Expert Systems with Applications · doi:10.1016/j.eswa.2021.114734 · Cited by 11
Intelligent mobile detection of cracks in concrete utilising an unmanned aerial vehicle
Alheeti, K. M. A., Al-Ani, M. S., Alaloosy, A. K. A. N., Alzahrani, A., & Rukan, D. A. S. (2022) · Bulletin of Electrical Engineering and Informatics · doi:10.11591/eei.v11i1.2987 · Cited by 10
“…problems such as in hypertext categorisation [12], image classification [13], hand-written recognition [14] face detection [15], object detection [16], fall detection [17], plant species classification [18], credit card fraud classification [19], bioinformatics [20], [21], intrusion detection in…”
BackgroundTextual description-based video summarization for video blogs
Otani, M., Nakashima, Y., Sato, T., & Yokoya, N. (2015) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2015.7177493 · Cited by 10
“, a method proposed in [21], can be applied.”
MethodAction recognition through discovering distinctive action parts
Chen, F., Sang, N., Kuang, X., Gan, H., & Gao, C. (2015) · Journal of the Optical Society of America A · doi:10.1364/josaa.32.000173 · Cited by 10
Interactive multiple object learning with scanty human supervision
Villamizar, M., Garrell, A., Sanfeliu, A., & Moreno-Noguer, F. (2016) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2016.03.010 · Cited by 11
“The standard method for recognizing objects in images consists in computing object specific classifiers during an offline and time-consuming training step where large amounts of annotated data are used to build discriminative and robust object detectors [15, 33].”
BackgroundMethodLiu, M., Zhang, K., Zhu, J., Wang, J., Guo, J. J., & Guo, Y. (2018) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2018.2880737 · Cited by 10
“For each patch, we train an Exemplar-SVM [40].”
MethodSemantic Cross-View MatchingPreprint
Castaldo, F., Zamir, A., Angst, R., Palmieri, F., & Savarese, S. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.00098 · Cited by 15
Montasari, R. (2024) · Advanced sciences and technologies for security applications · doi:10.1007/978-3-031-50454-9_10 · Cited by 10
Ensembles of exemplar-SVMs for video face recognition from a single sample per person
Bashbaghi, S., Granger, É., Sabourin, R., & Bilodeau, G.-A. (2015) · IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) · doi:10.1109/avss.2015.7301749 · Cited by 9
“In this paper, a multi-classifier system is proposed for robust still-to-video FR, where the single reference still of a target individual is modeled using an individual-specific ensemble of exemplar-SVM (e-SVM) classifiers [14].”
BackgroundMethodInfluentialAn automatic word-spotting framework for medieval manuscripts
Pintus, R., Yang, Y., Gobbetti, E., & Rushmeier, H. (2015) · 2015 Digital Heritage · doi:10.1109/digitalheritage.2015.7419446 · Cited by 9
Query Adaptive Similarity Measure for RGB-D Object Recognition
Cheng, Y., Cai, R., Zhang, C., Li, Z., Zhao, X., Huang, K., & Rui, Y. (2015) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.25 · Cited by 11
“An alternative way is to exploit the framework of exemplar-svm [27] while using our measure to evaluate the similarity between the pairwise objects.”
MethodGong, Y., Zhang, F., Jia, X., Mao, Z., Huang, X., & Li, D. (2021) · Remote Sensing · doi:10.3390/rs14010023 · Cited by 10
Learning with dataset bias in latent subcategory models
Stamos, D., Martelli, S., Nabi, M., McDonald, A., Murino, V., & Pontil, M. (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.2015.7298988 · Cited by 9
“Although our method tends to improve more over aggregated LSM than undoing bias SVMs, it is interesting to note that on the Caltech101 “person” or “dog” datasets, the trend reverses.”
BackgroundInfluentialPart and Attribute Discovery from Relative Annotations
Maji, S., & Shakhnarovich, G. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-014-0716-6 · Cited by 11
“Given a sampled ‘seed’, we initialize the model by training the HOG filter w(0) to separate the seed patch from a set of background patches; this step resembles the exemplar-SVM of Malisiewicz et al. (2011).”
MethodLearning Discriminative Collections of Part Detectors for Object Recognition
Shih, K. J., Endres, I., & Hoiem, D. (2014) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2014.2366122 · Cited by 10
Customized expression recognition for performance-driven cutout character animation
Yu, X., Yang, J., Luo, L., Li, W., Brandt, J., & Metaxas, D. (2016) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2016.7477449 · Cited by 9
“We repeat this procedure independently t times to produce tN classifiers, which we combine linearly to obtain the final ensemble classifier for expression i: FN (x) = tN∑ j=1 ωjfj(x), (6) where fj is the j-th SVM classifier trained using the positive samples and the j-th pile of negative samples, and ωj is its associa…”
BackgroundInfluentialContent-Based Cognitive Interference Control for City Monitoring Applications in the Urban IoT
Baidya, S., & Levorato, M. (2016) · Globecom · doi:10.1109/glocom.2016.7841693 · Cited by 9
“Popular object detection algorithms [19]–[21] include feature-based object detection, Viola-Jones object detection, object detection based on SVM (Support Vector Machines) and HOG (Histogram Oriented Gradient) features, Image segmentation and blob analysis.”
MethodWeak to Strong Detector Learning for Simultaneous Classification and Localization
Zhang, X., Xiong, H., Lin, W., & Tian, Q. (2018) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2018.2797923 · Cited by 9
“The weak detector learning module first selects patches which are representative and discriminative, then a series of exemplar-SVM [26] detectors are trained from each selected patch.”
BackgroundMethodInfluentialBERT-Based Meta-Learning Approach with Looking Back for Sentiment Analysis of Literary Book Reviews
Bao, H., He, K., Yin, X., Xuan-yu, L., Bao, X., Zhang, H., Wu, J., & Gao, Z. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-88483-3_18 · Cited by 9
Composite Discriminant Factor analysis
Morariu, V. I., Ahmed, E., Santhanam, V., Harwood, D., & Davis, L. S. (2014) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv.2014.6836052 · Cited by 10
“For instance, new sam- ples are efficiently projected into the subspace by matrix multiplication and the high-dimensional training data does not need to be stored as it is for kernel methods, reducing memory and computational requirements.”
BackgroundExploiting generic multi-level convolutional neural networks for scene understanding
Nguyen, T., Liu, L., & Nguyen, K. T. T. M. (2016) · International Conference on Control, Automation, Robotics and Vision (ICARCV) · doi:10.1109/icarcv.2016.7838719 · Cited by 9
“7 Our Method Settings Detector (reproduced [21]) 75.”
MethodSiqueira, H. L., Barros, P. V. A., Magg, S., & Wermter, S. (2018) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2018.8594276 · Cited by 11
“However, the majority of ensemble-based systems are composed of several independent classifiers, which leads to an unnecessary great redundancy of low-level features processing, and consequently, high computational cost [11].”
BackgroundMethodInfluentialAn online graph-based anomalous change detection strategy for unsupervised video surveillance
Kim, J., & Cho, J. (2019) · EURASIP Journal on Image and Video Processing · doi:10.1186/s13640-019-0478-8 · Cited by 9
Reweighting neural network examples for robust object detection at sea
Becktor, J., Boukas, E., Blanke, M., & Nalpantidis, L. (2021) · Electronics Letters · doi:10.1049/ell2.12166 · Cited by 9
“Hard example mining [8] down-samples the majority class and the hardest examples are exploited.”
BackgroundA hardware accelerated multilevel visual classifier for embedded visual-assist systems
Cotter, M. J., Advani, S., Sampson, J., Irick, K. M., & Narayanan, V. (2014) · International Conference on Computer Aided Design · doi:10.1109/iccad.2014.7001338 · Cited by 9
Compound Exemplar Based Object Detection by Incremental Random Forest
Ma, K., & Ben-Arie, J. (2014) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2014.417 · Cited by 9
“[23] proposed an exemplar-based SVM system that can be easily extended by training a few new exemplars with the negative samples.”
BackgroundMethodInfluentialSelf-Learning Camera: Autonomous Adaptation of Object Detectors to Unlabeled Video StreamsPreprint
Gaidon, A., Zen, G., & Rodríguez-Serrano, J. A. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1406.4296 · Cited by 9
“As we use linear classifiers, this multi-task learning could be seen as an improvement of “late fusion” of exemplar-based models, such as the Exemplar-SVM of Malisiewicz et al. [15].”
MethodWeakly Supervised One-Shot Detection with Attention Similarity NetworksPreprint
Keren, G., Schmitt, M., Kehrenberg, T., & Schuller, B. W. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1801.03329 · Cited by 11
Active Object Localization with Deep Reinforcement LearningPreprint
Caicedo, J. C., & Lazebnik, S. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1511.06015 · Cited by 28
“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].”
MethodExemplar-FreeSOLO: Enhancing Unsupervised Instance Segmentation with Exemplars
Ishtiak, T., En, Q., & Guo, Y. (2023) · 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/cvpr52729.2023.01480 · Cited by 14
“exemplar-based approach [30] has been proposed to address the object detection problem, but it requires supervised information to identify positive and negative samples, whereas our approach is unsupervised.”
MethodCombine color and shape in real-time detection of texture-less objects
Peng, X. (2015) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2015.02.010 · Cited by 10
Modality-Invariant Image Classification Based on Modality Uniqueness and Dictionary Learning
Kim, S., Cai, R., Park, K., Kim, S., & Sohn, K. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2635444 · Cited by 8
“It is derived from similar intuitions with an exemplar-SVM [52] and data-driven uniqueness [53], [54].”
MethodSpectral attribute learning for visual regression
Chen, K., Jia, K., Zhang, Z., & Kämäräinen, J. (2017) · Pattern Recognition · doi:10.1016/j.patcog.2017.01.009 · Cited by 9
“There are recent works on computational similarity functions by discriminative learning [24] and generative learning [21], but the computational (learning) similarity metrics require dedicated training process.”
Background
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.