Detection Results: VOC2010 BETA

Competition "comp3" (train on VOC2010 data)

This leaderboard shows only those submissions that have been marked as public, and so the displayed rankings should not be considered as definitive.Entries equivalent to a selected submission are determined by bootstrapping the performance measure, and assessing if the differences between the selected submission and the others are not statistically significant (see sec 3.5 in VOC 2014 paper).

Average Precision (AP %)

  mean

aero
plane
bicycle

bird

boat

bottle

bus

car

cat

chair

cow

dining
table
dog

horse

motor
bike
person

potted
plant
sheep

sofa

train

tv/
monitor
submission
date
NUS_UDS [?] 41.260.154.323.922.931.857.051.154.817.635.726.742.851.258.041.715.337.839.854.945.629-Oct-2014
Fisher with FLAIR [?] 40.461.352.327.825.721.354.045.654.015.532.633.341.847.957.837.324.341.835.850.447.317-Jun-2014
segDPM [?] 40.461.453.425.625.235.551.750.650.819.333.826.840.448.354.447.114.838.735.052.843.124-Feb-2014
Boosted HOG-LBP and multi-context (LC, EGC, HLC) [?] 36.853.355.319.221.030.054.546.741.220.031.520.830.348.655.346.510.234.426.650.340.329-Aug-2010
MITUCLA_Hierarchy [?] 36.054.348.515.719.229.255.643.541.716.928.526.730.948.355.041.79.735.830.847.240.830-Aug-2010
HOGLBP_context_classification_rescore_v2 [?] 34.249.152.417.812.030.653.532.837.317.730.627.729.551.956.344.29.614.827.949.538.430-Aug-2010
LSVM-MDPM [?] 33.752.454.313.015.635.154.249.131.815.526.213.521.545.451.647.59.135.119.446.638.026-Aug-2010
UOCTTI_LSVM_MDPM [?] 33.449.253.813.115.335.553.449.727.017.228.814.717.846.451.247.710.834.220.743.838.321-May-2012
Detection Monkey [?] 32.956.739.816.812.213.844.936.947.712.126.926.537.242.151.925.712.137.833.041.541.730-Aug-2010
RM^2C [?] 32.849.850.615.115.528.551.142.230.517.328.312.426.045.651.841.412.630.426.144.037.629-Oct-2013
UOCTTI_LSVM_MDPM [?] 32.248.252.214.813.828.753.244.926.018.424.413.723.145.850.543.79.831.121.544.435.711-May-2012
GroupLoc [?] 31.958.439.618.013.311.146.437.843.910.327.520.836.039.448.522.913.036.930.541.241.930-Aug-2010
UOCTTI_LSVM_MDPM [?] 29.645.649.011.011.627.250.543.123.617.223.210.720.542.544.541.38.729.018.740.034.521-May-2012
Bonn_FGT_Segm [?] 26.152.733.713.211.014.243.231.935.65.825.414.420.638.141.725.05.826.318.137.628.130-Aug-2010
HOG-LBP + DHOG bag of words, SVM [?] 23.540.434.72.78.426.043.133.817.211.214.314.514.931.837.330.06.425.211.630.035.730-Aug-2010
Svr-Segm [?] 23.450.524.517.113.310.939.532.936.55.616.06.622.324.929.029.86.728.413.332.127.230-Aug-2010
HOG-LBP Linear SVM [?] 22.137.933.72.76.525.337.533.115.510.912.312.513.729.734.533.87.222.99.928.934.129-Aug-2010
HOG+LBP+LTP+PLS2ROOTS [?] 17.532.729.70.81.119.939.427.58.64.58.16.311.022.934.124.63.124.02.023.527.031-Aug-2010
RandomParts [?] 14.223.831.71.23.411.129.719.514.20.811.17.04.716.431.516.01.115.610.214.721.025-Aug-2010
SIFT-GMM-MKL2 [?] 8.320.014.53.81.20.517.68.128.50.12.93.117.57.218.83.30.82.96.37.61.130-Aug-2010
UC3M_Generative_Discriminative [?] 6.315.85.55.62.30.310.25.412.60.55.64.57.711.312.65.31.52.05.99.13.230-Aug-2010
SIFT-GMM-MKL [?] 2.310.61.61.20.90.12.81.66.70.12.00.43.02.04.42.00.31.11.22.11.930-Aug-2010
CMIC_SynthTrain [?] --28.9---30.213.3-----26.228.113.2---18.825.730-Aug-2010
DPM-SP [?] -46.152.613.815.528.353.244.526.617.6-16.120.445.551.243.511.630.920.347.6-30-Aug-2010
RandomParts_maxScore [?] ---2.7----16.2-10.68.5---17.9---15.7-30-Aug-2010
CMIC_VarParts [?] --28.2---26.913.7-----23.524.716.1---18.824.530-Aug-2010

Abbreviations

TitleMethodAffiliationContributorsDescriptionDate
FG Detection, FG TilingBonn_FGT_SegmUniveristy of BonnJoão Carreira, Adrian Ion, Fuxin Li, Cristian Sminchisescu- Detection on figure-ground segmentations; - Classification by detection; - Semantic segmentation on figure-ground tilings;2010-08-30 23:08:18
NLPR_VSTAR_DET_4Boosted HOG-LBP and multi-context (LC, EGC, HLC)National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of SciencestntgroupDeformable model with Boosted HOG-LBP and multi-context information, use location context, enhanced global context, HOG and LBP inter-class context.2010-08-29 15:26:53
Synthetic Training of Deformable Part ModelsCMIC_SynthTrainCairo Microsoft Innovation Lab, Microsoft ResearchOsama Khalil, Yasmine Badr, Motaz El-SabanThis submission applies synthetic training for Deformable Part Models. Using the segmentation mask of the objects, we synthesized new training examples, by relocating the objects to different background. The idea was applied on top of the deformable models approach [1]. [1] P. Felzenszwalb, R. Girshick, D. McAllester, and D. Ramanan; Object detection with discriminatively trained part based models; PAMI 2009. 2010-08-30 21:16:16
Deformable part models with variable sized partsCMIC_VarPartsCairo Microsoft Innovation Lab, Microsoft ResearchOsama Khalil, Yasmine Badr, Motaz El-Saban Our submission is based on the Deformable Part Models approach[1]. We allowed the model parts to have variable sizes accommodating for affine distortion that renders part sizes non-proportional. [1] P. Felzenszwalb, R. Girshick, D. McAllester, and D. Ramanan; Object detection with discriminatively trained part based models; PAMI 2009. 2010-08-30 21:15:32
parts based model and spatial pyramid featuresDPM-SPUniversity of California, IrvineYi Yang, Charless FowlkesParts model results rescored by combining with spatial pyramid based scene (global) classification results. Scene trained using svm with hist intersect kernel, rescoring trained using SVM on train+val+some_from_segmentations with parameters obtained via search.2010-08-30 22:31:10
Detection MonkeyDetection MonkeyUniversity of AmsterdamKoen van de Sande Jasper Uijlings Theo Gevers Arnold SmeuldersThe detection monkey is trained with SVM, dense Color SIFT, spatial pyramid and multiple iterations.2010-08-30 20:46:02
Fisher with FLAIRFisher with FLAIRUniversity of AmsterdamKoen van de Sande, Cees Snoek, Arnold SmeuldersRun for our CVPR2014 paper "Fisher and VLAD with FLAIR", see http://koen.me/research/flair2014-06-17 11:41:40
Localisation with grouping window selectionGroupLocUniversity of AmsterdamJasper Uijlings Koen van de Sande Theo Gevers Arnold Smeulders Remko SchaCandidate windows are selected using hierarchical grouping. Classification is with SIFT, SVM-Histogram Intersection, Spatial Pyramid2010-08-30 21:58:30
HOG+LBP+LTP+PLS2ROOTSHOG+LBP+LTP+PLS2ROOTSLJK,INPGSibt ul Hussain, Bill TriggsThis method consists of two roots based object detector. This detector is also trained using HOG+LBP+LTP as feature sets while PLS based linear SVM is used as learning algorithm. 2010-08-31 01:14:00
Linear svm classifier with bag of words methodHOG-LBP + DHOG bag of words, SVMThe University of Missouri, NEC Labs America, The University of Illinois at Urbana-ChampaignXiaoyu Wang, Xi Zhou, Tony X. Han, Shuai Tang, Guang Chen, Kai Yu, Thomas S. HuangLiblinear SVM with HOG-LBP feature and DHOG bag of words approach2010-08-30 18:07:43
svm classifier with HOG LBP featuresHOG-LBP Linear SVMThe University of Missouri, NEC Labs America, The University of Illinois at Urbana-ChampaignXiaoyu Wang, Xi Zhou, Tony X. Han, Shuai Tang, Guang Chen, Kai Yu, Thomas S. HuangLiblinear SVM with HOG-LBP features. All classes use the same default training parameters. 2010-08-29 07:14:40
results refined by context and classificationHOGLBP_context_classification_rescore_v2National University of SingaporeZheng Song, Qiang Chen, Shuicheng YanUse HOG+LBP trained part-based detector. The detection results are further reranked via the context information of other detect windows and image classificaton scores.2010-08-30 04:20:46
LSVM Mixtures of deformable part modelsLSVM-MDPMUniversity of Chicago and TTI-CPedro Felzenszwalb (UChicago), Ross Girshick (UChicago), David McAllester (TTI-C)Our submission is based on an extension of [1,2]. Each model has 6 components with 8 parts. We associate a binary mask with each component to generate segmentations. The object detection models were trained from bounding boxes. The segmentation masks were trained from segmentations. [1] P. Felzenszwalb, R. Girshick, D. McAllester, D. Ramanan. Object Detection with Discriminatively Trained Part Based Models. PAMI 32(9), Sept 2010 [2] http://people.cs.uchicago.edu/~pff/latent-release4/ 2010-08-26 04:07:34
Latent hierarchical structural learningMITUCLA_HierarchyMIT and UCLALong Zhu, Yuanhao Chen, William Freeman, Alan Yuille, Antonio TorralbaLatent hierarchical structural learning with dense HOG and HOW(SIFT) features. 2010-08-30 20:58:06
Unified Object Detection and Semantic SegmentationNUS_UDSNUSJian Dong, Qiang Chen, Shuicheng Yan, Alan YuilleMotivated by the complementary effect observed from the typical failure cases of object detection and semantic segmentation, we propose a uni?ed framework for joint object detection and semantic segmentation [1]. By enforcing the consistency between final detection and segmentation results, our unified framework can effectively leverage the advantages of leading techniques for these two tasks. Furthermore, both local and global context information are integrated into the framework to better distinguish the ambiguous samples. By jointly optimizing the model parameters for all the components, the relative importance of different component is automatically learned for each category to guarantee the overall performance. [1] Jian Dong, Qiang Chen, Shuicheng Yan, Alan Yuille: Towards Unified Object Detection and Semantic Segmentation. ECCV 20142014-10-29 15:57:03
Randomized Max-Margin CompositionsRM^2CUniversity of Heidelberg, IWR/HCIAngela Eigenstetter, Masato Takami, Björn OmmerWe are grouping a large number of randomly sampled parts into fewer, overlapping compositions that are trained using a maximum-margin approach. For more details see our CVPR 2014 Paper "Randomized Max-Margin Compositions for Visual Recognition". Parts are available for download on our project page http://hci.iwr.uni-heidelberg.de/COMPVIS/research/RM2C/ . 2013-10-29 08:25:28
Unsupervised Parts-based AttributesRandomPartsCarnegie Mellon UniversitySantosh Divvala (CMU) Larry Zitnick (MSR) Ashish Kapoor (MSR) Simon Baker (MSR) http://www.cs.cmu.edu/~santosh/finalReport.pdf (unpublished work)2010-08-25 21:50:35
Unsupervised Parts-based Attributes (max score)RandomParts_maxScoreCarnegie Mellon UniversitySantosh Divvala (CMU) Larry Zitnick (MSR) Ashish Kapoor (MSR) Simon Baker (MSR) Updated version of earlier submission (http://www.cs.cmu.edu/~santosh/finalReport.pdf). Main update: inclusion of max score feature and one round of iterative training2010-08-30 23:10:27
Multiple kernel learning with SIFT GMMsSIFT-GMM-MKL Tokyo Institute of TechnologyNakamasa Inoue, Yusuke Kamishima, Koichi ShinodaWe use multiple kernel learning and GMM supervector kernels with SIFT features.2010-08-30 04:39:41
Multiple kernel learning with SIFT GMMsSIFT-GMM-MKL2Tokyo Institute of TechnologyNakamasa Inoue, Yusuke Kamishima, Koichi ShinodaSame as the SIFT-GMM-MKL run but the GrabCut is applied for detection.2010-08-30 07:35:56
Svr-SegmSvr-SegmUniversity of BonnJoao Carreira, Fuxin Li, Adrian Ion, Cristian SminchisescuSupport vector regression to multiple descriptors extracted on segmentations. Descriptors include SIFT, color SIFT and HOG on foreground and background. Post-processing to eliminate spurious detections and segmentations. The winning method of 2009 challenge.2010-08-30 22:52:55
Combination of Generative Discriminative MethodsUC3M_Generative_DiscriminativeUniversidad Carlos III de MadridIván González-Díaz, Fernando Díaz de MaríaCombination of Supervised Topic Models with SVM-based discriminative methods for concurrent image recognition and segmentation2010-08-30 13:27:41
LSVM mixtures of deformable part modelsUOCTTI_LSVM_MDPMUniversity of ChicagoRoss Girshickvoc-release52012-05-11 05:14:49
LSVM mixtures of deformable part modelsUOCTTI_LSVM_MDPMUniversity of ChicagoRoss Girshickvoc-release5 with extra octave and star-cascade2012-05-21 05:14:06
LSVM Mixtures of deformable part modelsUOCTTI_LSVM_MDPMUniversity of ChicagoRoss Girshickvoc-release5 without context 2012-05-21 17:39:28
DPM that uses region segmentation featuressegDPMUofT, TTI-C, UCLASanja Fidler, Roozbeh Mottaghi, Allan Yuille, Raquel UrtasunDPM-style model that exploits bottom-up segmentation (CPMC regions). In addition, we use context re-scoring based on object presence classifiers provided by NUS. Project page: http://www.cs.toronto.edu/~fidler/projects/segDPM.html2014-02-24 19:53:33