GSoC2016NmapSub/FPEngine.cc.diff at master · psdh/GSoC2016NmapSub · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
-#include "linear.h"
+#include <ml.h>
+
-static struct feature_node *vectorize(const FingerPrintResultsIPv6 *FPR) {
+static std::vector<float> vectorize(const FingerPrintResultsIPv6 *FPR, std::vector<float> features_vec) {
-  struct feature_node *features;
+
+  nr_feature = S1Model->getVarCount();
-  nr_feature = get_nr_feature(&FPModel);
-  features = new feature_node[nr_feature + 1];
-  for (i = 0; i < nr_feature; i++) {
-    features[i].index = i + 1;
-    features[i].value = -1;
-  }
-  features[i].index = -1;
+
-    features[idx++].value = vectorize_plen(resps[probe_name].getPacket());
-    features[idx++].value = vectorize_tc(resps[probe_name].getPacket());
-    features[idx++].value = vectorize_hlim(resps[probe_name].getPacket(), FPR->distance, FPR->distance_calculation_method);
+    features_vec[idx++] = vectorize_plen(resps[probe_name].getPacket());
+    features_vec[idx++] = vectorize_tc(resps[probe_name].getPacket());
+    features_vec[idx++] = vectorize_hlim(resps[probe_name].getPacket(), FPR->distance, FPR->distance_calculation_method);
-  features[idx++].value = vectorize_isr(resps);
+  features_vec[idx++] = vectorize_isr(resps);
-    features[idx++].value = tcp->getWindow();
+    features_vec[idx++] = tcp->getWindow();
-      features[idx++].value = (flags & mask) != 0;
+      features_vec[idx++] = (flags & mask) != 0;
-      features[idx++].value = opt.type;
+      features_vec[idx++] = opt.type;
-      features[idx++].value = opt.len;
+      features_vec[idx++] = opt.len;
-    features[idx++].value = mss;
-    features[idx++].value = sackok;
-    features[idx++].value = wscale;
+    features_vec[idx++] = mss;
+    features_vec[idx++] = sackok;
+    features_vec[idx++] = wscale;
-      features[idx++].value = (float)tcp->getWindow() / mss;
+      features_vec[idx++] = (float)tcp->getWindow() / mss;
-      features[idx++].value = -1;
+      features_vec[idx++] = -1;
-    features[idx++].value = vectorize_icmpv6_type(resps[probe_name].getPacket());
-    features[idx++].value = vectorize_icmpv6_code(resps[probe_name].getPacket());
+    features_vec[idx++] = vectorize_icmpv6_type(resps[probe_name].getPacket());
+    features_vec[idx++] = vectorize_icmpv6_code(resps[probe_name].getPacket());
-      log_write(LOG_PLAIN, "%.16g, ", features[i].value);
+      log_write(LOG_PLAIN, "%.16g, ", features_vec[i]);
-  return features;
+  return features_vec;
-  struct feature_node *features;
-  double *values;
-  struct label_prob *labels;
-
-  nr_class = get_nr_class(&FPModel);
-
-  features = vectorize(FPR);
-  values = new double[nr_class];
-  labels = new struct label_prob[nr_class];
-
-  apply_scale(features, get_nr_feature(&FPModel), FPscale);
-
-  predict_values(&FPModel, features, values);
-  for (i = 0; i < nr_class; i++) {
-    labels[i].label = i;
-    labels[i].prob = 1.0 / (1.0 + exp(-values[i]));
-  }
-  qsort(labels, nr_class, sizeof(labels[0]), label_prob_cmp);
-  for (i = 0; i < nr_class && i < MAX_FP_RESULTS; i++) {
-    FPR->matches[i] = &o.os_labels_ipv6[labels[i].label];
-    FPR->accuracy[i] = labels[i].prob;
-    FPR->num_matches = i + 1;
-    if (labels[i].prob >= 0.90 * labels[0].prob)
-      FPR->num_perfect_matches = i + 1;
-    if (o.debugging > 2) {
-      printf("%7.4f %7.4f %3u %s\n", FPR->accuracy[i] * 100,
-        novelty_of(features, labels[i].label), labels[i].label, FPR->matches[i]->OS_name);
-    }
-  }
-  if (FPR->num_perfect_matches == 0) {
-    FPR->overall_results = OSSCAN_NOMATCHES;
-  } else if (FPR->num_perfect_matches == 1) {
-    double novelty;
-
-    novelty = novelty_of(features, labels[0].label);
-    if (o.debugging > 1)
-      log_write(LOG_PLAIN, "Novelty of closest match is %.3f.\n", novelty);
-
-    if (novelty < FP_NOVELTY_THRESHOLD) {
-      FPR->overall_results = OSSCAN_SUCCESS;
-    } else {
-      if (o.debugging > 0) {
-        log_write(LOG_PLAIN, "Novelty of closest match is %.3f > %.3f; ignoring.\n",
-          novelty, FP_NOVELTY_THRESHOLD);
-      }
-      FPR->overall_results = OSSCAN_NOMATCHES;
-      FPR->num_perfect_matches = 0;
-    }
-  } else {
-    FPR->overall_results = OSSCAN_NOMATCHES;
-    FPR->num_perfect_matches = 0;
-  }
-
-  delete[] features;
-  delete[] values;
-  delete[] labels;
+  unsigned int nr_feature = S1Model->getVarCount();
+  std::vector<float> features(nr_feature, 0.0);
+
+  vectorize(FPR, features);
+
+  int s1Response = (int) S1Model->predict(features);
+
+  if (o.debugging > 1)
+    log_write(LOG_PLAIN, "Stage 1 prediction for MSRF is %d\n", s1Response);
+
+  // Loading the required stage 2 model
+  cv::Ptr<cv::ml::RTrees> stage2Model;
+  loadStage2Model(stage2Model, s1Response);
+
+  int s2Response = -1;
+
+  s2Response = (int) stage2Model->predict(features);
+
+  if (o.debugging > 1)
+    log_write(LOG_PLAIN, "Stage 2 prediction for MSRF is %d\n", s2Response);
+
+  std::cout<<"Stage 2 respnse"<<s2Response<<std::endl;