Skip to content

DwtDctSvdHideStrinWAV

stegobox.codec.DwtDctSvdHideStrinWAV

DwtDctSvd - embedding algorithm that creates stealthy steganographic audio and decodes them without relying on the original audio.

Progress:First,deal with audiio data. Read audio file in format bytes, and then transform them to bytearray(size:[N]) audio_data. Then process audio as a grayscale image,reshape the audio_data [:(a//block_size * block_size)*(b//block_size * block_size)] to size a//block_size * block_size,b//block_size * block_size dwt_data. Second, get the low-frequency part of the image after DWT transformation. and then use dct and svd to hide secret infomation in low-frequency.

Performance - DwtDctSvd is 3x slower than MaxDct.

Original implementation reports 1500ms-2s encoding time and ~1s decoding time for 1920x1080 images on CPU, not suitable for on-the-fly embedding for large images.

Originally implemented in ShieldMnt/invisible-watermark and is used inside CompVis/stable-diffusion.

Source code in stegobox/codec/dwtdctsvd_hide_str_in_wav.py 
 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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
class DwtDctSvdHideStrinWAV:
    """DwtDctSvd - embedding algorithm that creates stealthy steganographic audio and
    decodes them without relying on the original audio.

    Progress:First,deal with audiio data. Read audio file in format bytes, and then
    transform them to bytearray(size:[N]) audio_data. Then process audio as a grayscale
    image,reshape the audio_data
    [:(a//block_size * block_size)*(b//block_size * block_size)] to size
    [a//block_size * block_size,b//block_size * block_size](a*b=N) dwt_data.
        Second, get the low-frequency part of the image after DWT transformation.
    and then use dct and svd to hide secret infomation in low-frequency.

    Tip: Performance - DwtDctSvd is 3x slower than MaxDct.
        Original implementation reports 1500ms-2s encoding time and ~1s decoding time
        for 1920x1080 images on CPU, not suitable for on-the-fly embedding for large
        images.

    Originally implemented in
    [ShieldMnt/invisible-watermark](https://github.com/ShieldMnt/invisible-watermark)
    and is used inside
    [CompVis/stable-diffusion](https://github.com/CompVis/stable-diffusion).
    """

    def __init__(self, block: int = 4) -> None:
        """Frequency embedding method - DwtDctSvd uses DWT and DCT transforms, SVD
        decomposition of each block, to embed payload bits into singular value
        decomposition coefficients.

        Args:
            block: Number of DWT blocks. Defaults to 4.
        """
        self.block = block

    def payload_in_bytes(self, payload: str) -> list[np.uint8]:
        """
        transform the string into bytes,example 'A'->65

        Args:
            payload: _description_

        Returns:
            _description_
        """
        payloadb = payload.encode("utf-8")
        payloadsq = np.array([n for n in payloadb], dtype=np.uint8)
        return list(np.unpackbits(payloadsq))

    def reconstruct_bytes(self, bits: np.ndarray, length: int) -> str:
        bits_num = np.packbits(bits)
        bits_string = b""
        for i in range(length // 8):
            bits_string += struct.pack(">B", bits_num[i])
        return bits_string.decode("utf-8")

    def _spartnum(self, num: int) -> tuple[int, int]:
        """
        Break num into a product of two integers.
        example: num=a * b  the input num is 225
        the ouput is (15,15)
        Args:
            num: The integer to be decomposed.

        Returns:
            two integers and the product of them is num.
        """
        mid_num = int(num**0.5)
        a = mid_num
        b = int(num / a)
        if a * b == num:
            return a, b

        for i in range(mid_num, num, 1):
            a = a + 1
            b = int(num / a)
            if a * b == num:
                break
        return a, b

    def encode(
        self, carrier: wave.Wave_read, payload: str
    ) -> tuple[bytes, wave._wave_params, int]:
        """Encodes payload string into an audio with DWT + DCT transform and SVD.

        Args:
            carrier: Carrier audio in format WAV. Read with `stegobox.io.audio.read()`.
            payload: The payload string.

        Returns:
            The encoded wav audio.
            The headparams of audio.
            Length of payload in bits, you will need this when decoding.
        """
        payload_seq = self.payload_in_bytes(payload)
        payload_len = len(payload_seq)
        wave_head = carrier.getparams()
        audio_data = carrier.readframes(carrier.getnframes())
        audio_data = bytearray(audio_data)
        row, col = self._spartnum(len(audio_data))
        row = (row // self.block) * self.block
        col = (col // self.block) * self.block
        zong_length = row * col
        dct_data = audio_data[:zong_length]
        # deal the left data
        left_data = audio_data[zong_length:]

        # make the carrier data to N*4*2 block for dct's input
        dct_data = np.array(dct_data).reshape(row, col)
        dct_data = dct_data.astype(np.float32)

        # Payload string encoded as bytes
        print(f"Payload length: {payload_len} bits")

        c1, (h1, v1, d1) = pywt.dwt2(dct_data, "haar")
        self._encode_frame(payload_seq, payload_len, h1, 18)
        dct_data = pywt.idwt2((c1, (h1, v1, d1)), "haar")
        for i in range(len(dct_data)):
            for j in range(len(dct_data[i])):
                if dct_data[i][j] > 255:
                    dct_data[i][j] = 255

        dct_data = dct_data.reshape(-1).astype(np.uint8)

        outdata = []
        outdata.extend(dct_data)
        outdata.extend(left_data)
        # outdata=bytes(dct_data+left_data)
        outdata = bytes(outdata)
        return outdata, wave_head, payload_len

    def decode(self, _):
        raise NotImplementedError("This codec does not support decoding without length")

    def decode_with_length(self, carrier: wave.Wave_read, payload_len: int) -> str:
        """Try to decode payload from an audio with a DWT + DCT transform and SVD.

        Args:
            carrier: The encoded steganographic audio file.
            payload_len: The length of your payload in bits.

        Returns:
            The extracted payload string if successful.
        """
        audio_data = carrier.readframes(carrier.getnframes())
        audio_data = bytearray(audio_data)
        row, col = self._spartnum(len(audio_data))
        row = (row // self.block) * self.block
        col = (col // self.block) * self.block
        zong_length = row * col
        dct_data = audio_data[:zong_length]
        dct_data = np.array(dct_data).reshape(row, col)
        dct_data = dct_data.astype(np.float32)

        scores: list[list[int]] = [[] for _ in range(payload_len)]

        c1, (h1, _, _) = pywt.dwt2(dct_data, "haar")
        scores = self._decode_frame(payload_len, h1, 18, scores)
        avg_scores = list(map(lambda x: np.array(x).mean(), scores))
        bits = np.array(avg_scores) * 255 > 127
        return self.reconstruct_bytes(bits, payload_len)

    def _decode_frame(
        self, wm_len: int, frame: np.ndarray, scale: int, scores: list[list[int]]
    ) -> list[list[int]]:
        (row, col) = frame.shape
        num = 0

        for i in range(row // self.block):
            for j in range(col // self.block):
                block = frame[
                    i * self.block : i * self.block + self.block,
                    j * self.block : j * self.block + self.block,
                ]

                score = self._infer_dct_svd(block, scale)
                wm_bit = num % wm_len
                scores[wm_bit].append(score)
                num = num + 1

        return scores

    def _diffuse_dct_svd(
        self, block: np.ndarray, wm_bit: np.uint8, scale: int
    ) -> np.ndarray:
        u, s, v = np.linalg.svd(cv2.dct(block))

        s[0] = (s[0] // scale + 0.25 + 0.5 * wm_bit) * scale

        return cv2.idct(np.dot(u, np.dot(np.diag(s), v)))

    def _infer_dct_svd(self, block: np.ndarray, scale: int) -> int:
        u, s, v = np.linalg.svd(cv2.dct(block))

        score = 0
        score = int((s[0] % scale) > scale * 0.5)
        return score

    def _encode_frame(
        self, wm: list[np.uint8], wm_len: int, frame: np.ndarray, scale: int
    ) -> None:
        """
        frame is a matrix (M, N)

        we get K (watermark bits size) blocks (self.block x self.block)

        For i-th block, we encode watermark[i] bit into it
        """
        (row, col) = frame.shape
        num = 0
        for i in range(row // self.block):
            for j in range(col // self.block):
                block = frame[
                    i * self.block : i * self.block + self.block,
                    j * self.block : j * self.block + self.block,
                ]
                wm_bit = wm[(num % wm_len)]

                diffused_block = self._diffuse_dct_svd(block, wm_bit, scale)
                frame[
                    i * self.block : i * self.block + self.block,
                    j * self.block : j * self.block + self.block,
                ] = diffused_block

                num = num + 1

__init__(block=4)

Frequency embedding method - DwtDctSvd uses DWT and DCT transforms, SVD decomposition of each block, to embed payload bits into singular value decomposition coefficients.

Parameters:

Name Type Description Default
block int

Number of DWT blocks. Defaults to 4.

 4
Source code in stegobox/codec/dwtdctsvd_hide_str_in_wav.py 
34
35
36
37
38
39
40
41
42
def __init__(self, block: int = 4) -> None:
    """Frequency embedding method - DwtDctSvd uses DWT and DCT transforms, SVD
    decomposition of each block, to embed payload bits into singular value
    decomposition coefficients.

    Args:
        block: Number of DWT blocks. Defaults to 4.
    """
    self.block = block

payload_in_bytes(payload)

transform the string into bytes,example 'A'->65

Parameters:

Name Type Description Default
payload str

description

 required

Returns:

Type Description
list[uint8]

description
Source code in stegobox/codec/dwtdctsvd_hide_str_in_wav.py 
44
45
46
47
48
49
50
51
52
53
54
55
56
def payload_in_bytes(self, payload: str) -> list[np.uint8]:
    """
    transform the string into bytes,example 'A'->65

    Args:
        payload: _description_

    Returns:
        _description_
    """
    payloadb = payload.encode("utf-8")
    payloadsq = np.array([n for n in payloadb], dtype=np.uint8)
    return list(np.unpackbits(payloadsq))

encode(carrier, payload)

Encodes payload string into an audio with DWT + DCT transform and SVD.

Parameters:

Name Type Description Default
carrier Wave_read

Carrier audio in format WAV. Read with stegobox.io.audio.read().

 required
payload str

The payload string.

 required

Returns:

Type Description
bytes

The encoded wav audio.
_wave_params

The headparams of audio.
int

Length of payload in bits, you will need this when decoding.
Source code in stegobox/codec/dwtdctsvd_hide_str_in_wav.py 
 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
def encode(
    self, carrier: wave.Wave_read, payload: str
) -> tuple[bytes, wave._wave_params, int]:
    """Encodes payload string into an audio with DWT + DCT transform and SVD.

    Args:
        carrier: Carrier audio in format WAV. Read with `stegobox.io.audio.read()`.
        payload: The payload string.

    Returns:
        The encoded wav audio.
        The headparams of audio.
        Length of payload in bits, you will need this when decoding.
    """
    payload_seq = self.payload_in_bytes(payload)
    payload_len = len(payload_seq)
    wave_head = carrier.getparams()
    audio_data = carrier.readframes(carrier.getnframes())
    audio_data = bytearray(audio_data)
    row, col = self._spartnum(len(audio_data))
    row = (row // self.block) * self.block
    col = (col // self.block) * self.block
    zong_length = row * col
    dct_data = audio_data[:zong_length]
    # deal the left data
    left_data = audio_data[zong_length:]

    # make the carrier data to N*4*2 block for dct's input
    dct_data = np.array(dct_data).reshape(row, col)
    dct_data = dct_data.astype(np.float32)

    # Payload string encoded as bytes
    print(f"Payload length: {payload_len} bits")

    c1, (h1, v1, d1) = pywt.dwt2(dct_data, "haar")
    self._encode_frame(payload_seq, payload_len, h1, 18)
    dct_data = pywt.idwt2((c1, (h1, v1, d1)), "haar")
    for i in range(len(dct_data)):
        for j in range(len(dct_data[i])):
            if dct_data[i][j] > 255:
                dct_data[i][j] = 255

    dct_data = dct_data.reshape(-1).astype(np.uint8)

    outdata = []
    outdata.extend(dct_data)
    outdata.extend(left_data)
    # outdata=bytes(dct_data+left_data)
    outdata = bytes(outdata)
    return outdata, wave_head, payload_len

decode_with_length(carrier, payload_len)

Try to decode payload from an audio with a DWT + DCT transform and SVD.

Parameters:

Name Type Description Default
carrier Wave_read

The encoded steganographic audio file.

 required
payload_len int

The length of your payload in bits.

 required

Returns:

Type Description
str

The extracted payload string if successful.
Source code in stegobox/codec/dwtdctsvd_hide_str_in_wav.py 
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
def decode_with_length(self, carrier: wave.Wave_read, payload_len: int) -> str:
    """Try to decode payload from an audio with a DWT + DCT transform and SVD.

    Args:
        carrier: The encoded steganographic audio file.
        payload_len: The length of your payload in bits.

    Returns:
        The extracted payload string if successful.
    """
    audio_data = carrier.readframes(carrier.getnframes())
    audio_data = bytearray(audio_data)
    row, col = self._spartnum(len(audio_data))
    row = (row // self.block) * self.block
    col = (col // self.block) * self.block
    zong_length = row * col
    dct_data = audio_data[:zong_length]
    dct_data = np.array(dct_data).reshape(row, col)
    dct_data = dct_data.astype(np.float32)

    scores: list[list[int]] = [[] for _ in range(payload_len)]

    c1, (h1, _, _) = pywt.dwt2(dct_data, "haar")
    scores = self._decode_frame(payload_len, h1, 18, scores)
    avg_scores = list(map(lambda x: np.array(x).mean(), scores))
    bits = np.array(avg_scores) * 255 > 127
    return self.reconstruct_bytes(bits, payload_len)