Skip to content

A
audiowmark
Commit 9516c0d6 authored by Stefan Westerfeld's avatar Stefan Westerfeld
Browse files
Options

Implemented non-blind decoder. 

Signed-off-by: Stefan Westerfeld's avatarStefan Westerfeld <stefan@space.twc.de>
parent db63439d
Hide whitespace changes
Inline Side-by-side
Showing with 161 additions and 67 deletions
src/TODO
View file @ 9516c0d6
- non-blind decoder
- standard tests (-> mp3, ->ogg, ...)
- peaq scores
src/audiowmark.cc
View file @ 9516c0d6
......@@ -146,6 +146,81 @@ bit_vec_to_str (const vector<int>& bit_vec)
return bit_str;
}
vector<float>
watermark_frame (const vector<float>& input_frame, int f, int data_bit)
{
vector<float> frame = input_frame;
/* windowing */
double window_weight = 0;
for (size_t i = 0; i < frame.size(); i++)
{
const double fsize_2 = frame.size() / 2.0;
// const double win = window_cos ((i - fsize_2) / fsize_2);
const double win = window_hamming ((i - fsize_2) / fsize_2);
//const double win = 1;
frame[i] *= win;
window_weight += win;
}
/* to get normalized fft output corrected by window weight */
for (size_t i = 0; i < frame.size(); i++)
frame[i] *= 2.0 / window_weight;
/* FFT transform */
vector<complex<float>> fft_out = fft (frame);
vector<complex<float>> fft_delta_spect (fft_out.size());
vector<int> up;
vector<int> down;
get_up_down (f, up, down);
const double data_bit_sign = data_bit > 0 ? 1 : -1;
for (auto u : up)
{
/*
* for up bands, we want do use [for a 1 bit] (pow (mag, 1 - water_delta))
*
* this actually increases the amount of energy because mag is less than 1.0
*/
const float mag_factor = pow (abs (fft_out[u]), -Params::water_delta * data_bit_sign);
fft_delta_spect[u] = fft_out[u] * (mag_factor - 1);
}
for (auto d : down)
{
/*
* for down bands, we want do use [for a 1 bit] (pow (mag, 1 + water_delta))
*
* this actually decreases the amount of energy because mag is less than 1.0
*/
const float mag_factor = pow (abs (fft_out[d]), Params::water_delta * data_bit_sign);
fft_delta_spect[d] = fft_out[d] * (mag_factor - 1);
}
/* add watermark to output frame */
vector<float> fft_delta_out = ifft (fft_delta_spect);
vector<float> synth_window (Params::frame_size);
for (int i = 0; i < Params::frame_size; i++)
{
const double threshold = 0.2;
// triangular basic window
const double tri = min (1.0 - fabs (double (2 * i)/Params::frame_size - 1.0), threshold) / threshold;
// cosine
synth_window[i] = (cos (tri*M_PI+M_PI)+1) * 0.5;
}
vector<float> new_frame = input_frame;
for (size_t i = 0; i < frame.size(); i++)
new_frame[i] += fft_delta_out[i] * synth_window[i];
return new_frame;
}
int
add_watermark (const string& infile, const string& outfile, const string& bits)
{
......@@ -167,80 +242,22 @@ add_watermark (const string& infile, const string& outfile, const string& bits)
vector<float> out_signal (wav_data.n_values());
printf ("channels: %d, samples: %zd, mix_freq: %f\n", wav_data.n_channels(), wav_data.n_values(), wav_data.mix_freq());
vector<float> synth_window (Params::frame_size);
for (int i = 0; i < Params::frame_size; i++)
{
const double threshold = 0.2;
// triangular basic window
const double tri = min (1.0 - fabs (double (2 * i)/Params::frame_size - 1.0), threshold) / threshold;
// cosine
synth_window[i] = (cos (tri*M_PI+M_PI)+1) * 0.5;
}
for (int f = 0; f < frame_count (wav_data); f++)
{
for (int ch = 0; ch < wav_data.n_channels(); ch++)
{
vector<float> frame = get_frame (wav_data, f, ch);
vector<float> new_frame = frame; /* will be modified below */
vector<float> new_frame;
if (frame.size() == Params::frame_size)
{
/* windowing */
double window_weight = 0;
for (size_t i = 0; i < frame.size(); i++)
{
const double fsize_2 = frame.size() / 2.0;
// const double win = window_cos ((i - fsize_2) / fsize_2);
const double win = window_hamming ((i - fsize_2) / fsize_2);
//const double win = 1;
frame[i] *= win;
window_weight += win;
}
/* to get normalized fft output corrected by window weight */
for (size_t i = 0; i < frame.size(); i++)
frame[i] *= 2.0 / window_weight;
/* FFT transform */
vector<complex<float>> fft_out = fft (frame);
vector<complex<float>> fft_delta_spect (fft_out.size());
vector<int> up;
vector<int> down;
get_up_down (f, up, down);
const int data_bit = bitvec[(f / Params::frames_per_bit) % bitvec.size()];
const int data_bit = bitvec[(f / Params::frames_per_bit) % bitvec.size()];
const double data_bit_sign = data_bit > 0 ? 1 : -1;
for (auto u : up)
{
/*
* for up bands, we want do use [for a 1 bit] (pow (mag, 1 - water_delta))
*
* this actually increases the amount of energy because mag is less than 1.0
*/
const float mag_factor = pow (abs (fft_out[u]), -Params::water_delta * data_bit_sign);
fft_delta_spect[u] = fft_out[u] * (mag_factor - 1);
}
for (auto d : down)
{
/*
* for down bands, we want do use [for a 1 bit] (pow (mag, 1 + water_delta))
*
* this actually decreases the amount of energy because mag is less than 1.0
*/
const float mag_factor = pow (abs (fft_out[d]), Params::water_delta * data_bit_sign);
fft_delta_spect[d] = fft_out[d] * (mag_factor - 1);
}
/* add watermark to output frame */
vector<float> fft_delta_out = ifft (fft_delta_spect);
for (size_t i = 0; i < new_frame.size(); i++)
new_frame[i] += fft_delta_out[i] * synth_window[i];
new_frame = watermark_frame (frame, f, data_bit);
}
else
{
new_frame = frame;
}
for (size_t i = 0; i < new_frame.size(); i++)
out_signal[(f * Params::frame_size + i) * wav_data.n_channels() + ch] = new_frame[i] * Params::pre_scale;
......@@ -338,6 +355,76 @@ get_watermark (const string& origfile, const string& infile, const string& orig_
return 0;
}
int
get_watermark_delta (const string& origfile, const string& infile, const string& orig_pattern)
{
WavData orig_wav_data;
if (!orig_wav_data.load (origfile))
{
fprintf (stderr, "audiowmark: error loading %s: %s\n", origfile.c_str(), orig_wav_data.error_blurb());
return 1;
}
WavData wav_data;
if (!wav_data.load (infile))
{
fprintf (stderr, "audiowmark: error loading %s: %s\n", infile.c_str(), wav_data.error_blurb());
return 1;
}
vector<int> bit_vec;
double error0 = 0;
double error1 = 0;
for (int f = 0; f < frame_count (wav_data); f++)
{
for (int ch = 0; ch < wav_data.n_channels(); ch++)
{
/* prescale original data (may want to do energy normalization or similar instead) */
vector<float> orig_frame = get_frame (orig_wav_data, f, ch);
for (auto& orig_value : orig_frame)
orig_value *= Params::pre_scale;
vector<float> frame = get_frame (wav_data, f, ch);
if (frame.size() == Params::frame_size)
{
vector<float> f0 = watermark_frame (orig_frame, f, 0);
vector<float> f1 = watermark_frame (orig_frame, f, 1);
for (size_t i = 0; i < frame.size(); i++)
{
const double delta0 = frame[i] - f0[i];
error0 += delta0 * delta0;
const double delta1 = frame[i] - f1[i];
error1 += delta1 * delta1;
}
}
}
if ((f % Params::frames_per_bit) == (Params::frames_per_bit - 1))
{
bit_vec.push_back ((error0 > error1) ? 1 : 0);
error0 = 0;
error1 = 0;
}
}
printf ("pattern %s\n", bit_vec_to_str (bit_vec).c_str());
if (!orig_pattern.empty())
{
int bits = 0, bit_errors = 0;
vector<int> orig_vec = bit_str_to_vec (orig_pattern);
for (size_t i = 0; i < bit_vec.size(); i++)
{
bits++;
if (bit_vec[i] != orig_vec[i % orig_vec.size()])
bit_errors++;
}
printf ("bit_error_rate %.5f %%\n", double (100.0 * bit_errors) / bits);
}
return 0;
}
int
gentest (const string& infile, const string& outfile)
{
......@@ -439,6 +526,14 @@ main (int argc, char **argv)
{
get_snr (argv[2], argv[3]);
}
else if (op == "get-delta" && argc == 4)
{
return get_watermark_delta (argv[2], argv[3], /* no ber */ "");
}
else if (op == "cmp-delta" && argc == 5)
{
return get_watermark_delta (argv[2], argv[3], argv[4]);
}
else
{
fprintf (stderr, "audiowmark: error parsing commandline args\n");
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment