File size: 5,114 Bytes
b92918a
 
 
 
 
 
 
 
 
 
52b5518
b92918a
 
 
 
 
 
 
 
 
 
 
 
 
 
52b5518
 
 
 
 
b92918a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52b5518
 
 
 
 
b92918a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
import os
import sys
import argparse
import numpy as np
import tensorflow as tf
import pandas as pd

sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

from src.utils import read_binary_file
from src.model import MalConv

def predict_file(model_path, file_path, max_length=2_000_000):  # 2,000,000
    """
    ๋‹จ์ผ ํŒŒ์ผ์— ๋Œ€ํ•œ ์˜ˆ์ธก
    
    Args:
        model_path: ์ €์žฅ๋œ ๋ชจ๋ธ ๊ฒฝ๋กœ
        file_path: ์˜ˆ์ธกํ•  ํŒŒ์ผ ๊ฒฝ๋กœ
        max_length: ์ตœ๋Œ€ ์ž…๋ ฅ ๊ธธ์ด
    
    Returns:
        float: ์˜ˆ์ธก ํ™•๋ฅ  (0์— ๊ฐ€๊นŒ์šฐ๋ฉด ์•…์„ฑ์ฝ”๋“œ, 1์— ๊ฐ€๊นŒ์šฐ๋ฉด ์ •์ƒ)
    """
    # ๋ชจ๋ธ ๋กœ๋“œ
    model = MalConv(max_input_length=max_length)
    # ๋ชจ๋ธ์˜ ๊ฐ€์ค‘์น˜๋ฅผ ๋กœ๋“œํ•˜๊ธฐ ์ „์— ๋นŒ๋“œ
    dummy_input = tf.zeros((1, max_length), dtype=tf.int32)
    model(dummy_input)  # ๋ชจ๋ธ ๋นŒ๋“œ
    model.load_weights(model_path)
    
    # ํŒŒ์ผ ์ฝ๊ธฐ
    byte_array = read_binary_file(file_path, max_length)
    
    # ๋ฐฐ์น˜ ์ฐจ์› ์ถ”๊ฐ€
    input_data = np.expand_dims(byte_array, axis=0)
    
    # ์˜ˆ์ธก
    prediction = model.predict(input_data, verbose=0)[0][0]
    
    return prediction

def predict_batch(model_path, csv_path, output_path=None, max_length=2**20):
    """
    ๋ฐฐ์น˜ ์˜ˆ์ธก
    
    Args:
        model_path: ์ €์žฅ๋œ ๋ชจ๋ธ ๊ฒฝ๋กœ
        csv_path: ์˜ˆ์ธกํ•  ํŒŒ์ผ๋“ค์˜ CSV ๊ฒฝ๋กœ
        output_path: ๊ฒฐ๊ณผ ์ €์žฅ ๊ฒฝ๋กœ
        max_length: ์ตœ๋Œ€ ์ž…๋ ฅ ๊ธธ์ด
    """
    # ๋ชจ๋ธ ๋กœ๋“œ
    print("๋ชจ๋ธ ๋กœ๋”ฉ ์ค‘...")
    model = MalConv(max_input_length=max_length)
    # ๋ชจ๋ธ์˜ ๊ฐ€์ค‘์น˜๋ฅผ ๋กœ๋“œํ•˜๊ธฐ ์ „์— ๋นŒ๋“œ
    dummy_input = tf.zeros((1, max_length), dtype=tf.int32)
    model(dummy_input)  # ๋ชจ๋ธ ๋นŒ๋“œ
    model.load_weights(model_path)
    
    # CSV ํŒŒ์ผ ์ฝ๊ธฐ
    df = pd.read_csv(csv_path)
    
    predictions = []
    labels = []
    
    print("์˜ˆ์ธก ์ค‘...")
    for idx, row in df.iterrows():
        file_path = row['filepath']
        
        if os.path.exists(file_path):
            try:
                # ํŒŒ์ผ ์ฝ๊ธฐ
                byte_array = read_binary_file(file_path, max_length)
                input_data = np.expand_dims(byte_array, axis=0)
                
                # ์˜ˆ์ธก
                pred = model.predict(input_data, verbose=0)[0][0]
                predictions.append(pred)
                
                # ๋ผ๋ฒจ์ด ์žˆ๋Š” ๊ฒฝ์šฐ
                if 'label' in row:
                    labels.append(row['label'])
                
                # ๊ฒฐ๊ณผ ์ถœ๋ ฅ
                status = "์ •์ƒ" if pred > 0.5 else "์•…์„ฑ์ฝ”๋“œ"
                confidence = pred if pred > 0.5 else 1 - pred
                print(f"{file_path}: {status} (์‹ ๋ขฐ๋„: {confidence:.4f})")
                
            except Exception as e:
                print(f"Error processing {file_path}: {e}")
                predictions.append(-1)  # ์—๋Ÿฌ ํ‘œ์‹œ
        else:
            print(f"ํŒŒ์ผ์„ ์ฐพ์„ ์ˆ˜ ์—†์Šต๋‹ˆ๋‹ค: {file_path}")
            predictions.append(-1)
    
    # ๊ฒฐ๊ณผ ์ €์žฅ
    result_df = df.copy()
    result_df['prediction'] = predictions
    result_df['predicted_label'] = (np.array(predictions) > 0.5).astype(int)
    result_df['prediction_text'] = ['์ •์ƒ' if p > 0.5 else '์•…์„ฑ์ฝ”๋“œ' if p >= 0 else '์—๋Ÿฌ' 
                                   for p in predictions]
    
    if output_path:
        result_df.to_csv(output_path, index=False)
        print(f"๊ฒฐ๊ณผ๊ฐ€ ์ €์žฅ๋˜์—ˆ์Šต๋‹ˆ๋‹ค: {output_path}")
    
    # ์ •ํ™•๋„ ๊ณ„์‚ฐ (๋ผ๋ฒจ์ด ์žˆ๋Š” ๊ฒฝ์šฐ)
    if labels and len(labels) == len(predictions):
        valid_predictions = [p for p in predictions if p >= 0]
        valid_labels = [labels[i] for i, p in enumerate(predictions) if p >= 0]
        
        if valid_predictions:
            pred_binary = (np.array(valid_predictions) > 0.5).astype(int)
            accuracy = np.mean(pred_binary == np.array(valid_labels))
            print(f"\n์ •ํ™•๋„: {accuracy:.4f}")
    
    return result_df

def main():
    parser = argparse.ArgumentParser(description='MalConv ๋ชจ๋ธ ์˜ˆ์ธก')
    parser.add_argument('model_path', help='์ €์žฅ๋œ ๋ชจ๋ธ ๊ฒฝ๋กœ')
    parser.add_argument('--file', help='๋‹จ์ผ ํŒŒ์ผ ์˜ˆ์ธก')
    parser.add_argument('--csv', help='๋ฐฐ์น˜ ์˜ˆ์ธก์šฉ CSV ํŒŒ์ผ')
    parser.add_argument('--output', help='๊ฒฐ๊ณผ ์ €์žฅ ๊ฒฝ๋กœ')
    parser.add_argument('--max_length', type=int, default=2**20, help='์ตœ๋Œ€ ์ž…๋ ฅ ๊ธธ์ด')
    
    args = parser.parse_args()
    
    if args.file:
        # ๋‹จ์ผ ํŒŒ์ผ ์˜ˆ์ธก
        prediction = predict_file(args.model_path, args.file, args.max_length)
        status = "์ •์ƒ" if prediction > 0.5 else "์•…์„ฑ์ฝ”๋“œ"
        confidence = prediction if prediction > 0.5 else 1 - prediction
        print(f"ํŒŒ์ผ: {args.file}")
        print(f"์˜ˆ์ธก: {status} (์‹ ๋ขฐ๋„: {confidence:.4f})")
        
    elif args.csv:
        # ๋ฐฐ์น˜ ์˜ˆ์ธก
        predict_batch(args.model_path, args.csv, args.output, args.max_length)
        
    else:
        print("--file ๋˜๋Š” --csv ์˜ต์…˜์„ ์ง€์ •ํ•ด์ฃผ์„ธ์š”.")

if __name__ == "__main__":
    main()