EURASIP Journal on Image and Video Processing

Table 8 Event classification rate (%) in UIUC Sports Event

From: DTCTH: a discriminative local pattern descriptor for image classification

Techniques	Accuracy
Places-CNN, 2014 [71]	94.12
ImageNet-CNN, 2014 [71]	94.42
Hybride-CNN, 2014 [71]	94.22
Dense SIFT (KSRSPM) LSVM, 2010 [65]	84.92±0.78
Dense SIFT (ScSPM + MP) LSVM,
2009 [60]	82.74±1.46
Dense SIFT (LSA + MMP) LSVM,
2011 [68]	82.29±1.84
Dense SIFT (LLC + MP) LSVM,
2010 [62]	81.41±1.84
Dense SIFT (LCSR + MP) LSVM,
2012 [61]	87.23±1.14
Dense SIFT + I2CDML, 2010 [58]	78.5±1.63
Dense SIFT (SPM) I2CDML, 2010 [58]	79.7±1.83
Dense SIFT + NBNN, 2008 [58, 59]	67.6±1.1
Dense SIFT (BoF) SVM with HI,
2014 [30, 93]	85.12
LQP + SVM with RBF, 2012 [39, 51]	78.9
DDLBP + Max Relevance +
SVM with RBF, 2014 [39]	83.5
DDLBP + mRMR + SVM with RBF,
2014 [39]	83.5
DDLBP + MJMI + SVM with RBF,
2014 [39]	84.0
mGIST (SPM) LSVM, 2014 [6]	76.2±1.9
mSIFT (SPM) LSVM, 2014 [6]	84.2±0.7
mCENTRIST (SPM) LSVM, 2014 [6]	86.5±0.6
LGP (SPM) LSVM, 2013	78.42±0.94
OC-LBP (BoF) LSVM, 2013	81.15±2.18
LAID (SPM) LSVM, 2013	78.50±0.65
CLBP_S/M/C (SPM) LSVM, 2010	78.88±0.92
LTP (SPM) LSVM, 2010	82.43±1.17
GIST + LSVM, 2001	69.95±0.98
CENTRIST (SPM) LSVM, 2011	79.50±0.95
Proposed (DTCTH + LSVM)	85.16±0.96
Proposed (DTCTH + HI)	88.18±0.84

Back to article page