diff --git a/openbr/plugins/liblinear.cmake b/openbr/plugins/liblinear.cmake
new file mode 100644
index 0000000..3845f08
--- /dev/null
+++ b/openbr/plugins/liblinear.cmake
@@ -0,0 +1,7 @@
+set(BR_WITH_LIBLINEAR OFF CACHE BOOL "Build with LibLinear")
+
+if(${BR_WITH_LIBLINEAR})
+  find_package(LibLinear REQUIRED)
+  set(BR_THIRDPARTY_SRC ${BR_THIRDPARTY_SRC} ${LibLinear_SRC})
+  set(BR_THIRDPARTY_SRC ${BR_THIRDPARTY_SRC} plugins/liblinear.cpp)
+endif()
diff --git a/openbr/plugins/liblinear.cpp b/openbr/plugins/liblinear.cpp
new file mode 100644
index 0000000..4ecf72c
--- /dev/null
+++ b/openbr/plugins/liblinear.cpp
@@ -0,0 +1,147 @@
+#include <QTemporaryFile>
+#include <opencv2/core/core.hpp>
+#include <opencv2/ml/ml.hpp>
+
+#include "openbr_internal.h"
+#include "openbr/core/opencvutils.h"
+
+#include <linear.h>
+#define Malloc(type,n) (type *)malloc((n)*sizeof(type))
+
+using namespace cv;
+
+namespace br
+{
+
+class LinearSVM : public Transform
+{
+    Q_OBJECT
+    Q_ENUMS(Kernel)
+    Q_ENUMS(Type)
+    Q_PROPERTY(Kernel kernel READ get_kernel WRITE set_kernel RESET reset_kernel STORED false)
+    Q_PROPERTY(Type type READ get_type WRITE set_type RESET reset_type STORED false)
+    Q_PROPERTY(float C READ get_C WRITE set_C RESET reset_C STORED false)
+    Q_PROPERTY(float gamma READ get_gamma WRITE set_gamma RESET reset_gamma STORED false)
+    Q_PROPERTY(QString inputVariable READ get_inputVariable WRITE set_inputVariable RESET reset_inputVariable STORED false)
+    Q_PROPERTY(QString outputVariable READ get_outputVariable WRITE set_outputVariable RESET reset_outputVariable STORED false)
+    Q_PROPERTY(bool returnDFVal READ get_returnDFVal WRITE set_returnDFVal RESET reset_returnDFVal STORED false)
+    Q_PROPERTY(int termCriteria READ get_termCriteria WRITE set_termCriteria RESET reset_termCriteria STORED false)
+    Q_PROPERTY(int folds READ get_folds WRITE set_folds RESET reset_folds STORED false)
+    Q_PROPERTY(bool balanceFolds READ get_balanceFolds WRITE set_balanceFolds RESET reset_balanceFolds STORED false)
+
+public:
+    enum Kernel { Linear = CvSVM::LINEAR,
+                  Poly = CvSVM::POLY,
+                  RBF = CvSVM::RBF,
+                  Sigmoid = CvSVM::SIGMOID };
+
+    enum Type { C_SVC = CvSVM::C_SVC,
+                NU_SVC = CvSVM::NU_SVC,
+                ONE_CLASS = CvSVM::ONE_CLASS,
+                EPS_SVR = CvSVM::EPS_SVR,
+                NU_SVR = CvSVM::NU_SVR};
+
+private:
+    BR_PROPERTY(Kernel, kernel, Linear)
+    BR_PROPERTY(Type, type, C_SVC)
+    BR_PROPERTY(float, C, -1)
+    BR_PROPERTY(float, gamma, -1)
+    BR_PROPERTY(QString, inputVariable, "Label")
+    BR_PROPERTY(QString, outputVariable, "")
+    BR_PROPERTY(bool, returnDFVal, false)
+    BR_PROPERTY(int, termCriteria, 1000)
+    BR_PROPERTY(int, folds, 5)
+    BR_PROPERTY(bool, balanceFolds, false)
+
+    model *m;
+
+    void train(const TemplateList &data)
+    {
+        Mat samples = OpenCVUtils::toMat(data.data());
+        Mat labels = OpenCVUtils::toMat(File::get<float>(data, inputVariable));
+
+        // Number of features = n
+        // Number of instances = l
+
+        problem prob;
+        prob.n = samples.cols;
+        prob.l = samples.rows;
+        prob.bias = -1;
+        prob.y = new double[prob.l];
+
+        for (int i=0; i<prob.l; i++)
+            prob.y[i] = labels.at<float>(i,0);
+
+        // Allocate enough memory for l feature_nodes pointers
+        prob.x = new feature_node*[prob.l];
+        feature_node *x_space = new feature_node[(prob.n+1)*prob.l];
+
+        int k = 0;
+        for (int i=0; i<prob.l; i++) {
+            prob.x[i] = &x_space[k];
+            for (int j=0; j<prob.n; j++) {
+                x_space[k].index = j+1;
+                x_space[k].value = samples.at<float>(i,j);
+                k++;
+            }
+            x_space[k++].index = -1;
+        }
+
+        parameter param;
+        param.C = 1;
+        param.eps = FLT_EPSILON;
+        param.solver_type = L2R_L2LOSS_SVC_DUAL;
+        param.nr_weight = 0;
+        param.p = 1;
+        param.weight_label = NULL;
+        param.weight = NULL;
+
+        m = train_svm(&prob, &param);
+
+        delete x_space;
+        delete prob.x;
+        delete prob.y;
+    }
+
+    void project(const Template &src, Template &dst) const
+    {
+        dst = src;
+
+        Mat sample = src.m().reshape(1,1);
+        feature_node *x_space = new feature_node[sample.cols];
+
+        // Assign the address of the ith instance to be the address of the jth feature
+        for (int j=0; j<sample.cols; j++) {
+            x_space[j].index = j+1;
+            x_space[j].value = sample.at<float>(0,j);
+        }
+        x_space[sample.cols].index = -1;
+
+        double prob_estimates[1];
+        float prediction = predict_values(m,x_space,prob_estimates);
+
+        delete x_space;
+        dst.m() = Mat(1, 1, CV_32F);
+        dst.m().at<float>(0, 0) = prob_estimates[0];
+    }
+
+    void store(QDataStream &stream) const
+    {
+        QString filename = QString::number(qrand());
+        stream << filename;
+        save_model(filename.toStdString().c_str(),m);
+    }
+
+    void load(QDataStream &stream)
+    {
+        QString filename;
+        stream >> filename;
+        m = load_model(filename.toStdString().c_str());
+    }
+};
+
+BR_REGISTER(Transform, LinearSVM)
+
+} // namespace br
+
+#include "liblinear.moc"
diff --git a/share/openbr/cmake/FindLibLinear.cmake b/share/openbr/cmake/FindLibLinear.cmake
new file mode 100644
index 0000000..795f4dc
--- /dev/null
+++ b/share/openbr/cmake/FindLibLinear.cmake
@@ -0,0 +1,13 @@
+find_path(LibLinear_DIR linear.h ${CMAKE_SOURCE_DIR}/3rdparty/*)
+
+message(${LibLinear_DIR})
+mark_as_advanced(LibLinear_DIR)
+include_directories(${LibLinear_DIR})
+include_directories(${LibLinear_DIR}/blas)
+
+set(LibLinear_SRC ${LibLinear_DIR}/linear.cpp 
+					${LibLinear_DIR}/tron.cpp
+					${LibLinear_DIR}/blas/daxpy.c
+					${LibLinear_DIR}/blas/ddot.c
+					${LibLinear_DIR}/blas/dnrm2.c
+					${LibLinear_DIR}/blas/dscal.c)