Research Article

Optimization of a Blind Speech Watermarking Technique against Amplitude Scaling

Table 2

Comparison with the robustness of different speech watermarking techniques in terms of BER (%).

AttackThe proposed methodDWPT+ multiplication [14]Formant tuning [21]Analysis-by-Synthesis [22]

No attack0.000.000.040.06
A1.91–4.232.09–5.433.65–6.457.96–9.65
B9.65–21.7710.45–22.3212.76–24.4516.23–25.23
C10.13–20.4312.43–21.3214.23–23.5417.43–26.32
D10.53–19.2310.33–18.9311.63–23.2315.33–25.98
E 0.32–2.020.763–1.141.23–2.322.98–4.32
F13.54–17.2314.32–17.6526.23–37.8329.45–33.06
G3.232.6519.3223.87
H0.230.009.4312.45
I1.34–4.652.34–5.114.65–10.438.23–16.43
J1.23–2.541.32–4.676.54–10.5411.54–18.87
K1.32–3.161.78–4.237.51–10.3411.49–19.43
L 0.921.981.504.04
M3.125.7610.3421.68
N4.104.236.659.54
O1.21–2.540.00–1.435.97–8.768.98–15.54
P1.00–3.542.43–5.439.65–14.5619.65–26.45
Q21.43–29.4324.54–31.4340.54–44.4350.09–50.32
R4.84–9.545.32–10.3216.65–29.4420.54–36.98
S13.32–18.5415.00–19.4320.43–29.2328.54–30.76
T1.32–2.322.01–3.137.43–10.439.65–15.32
U0.15–0.230.18–0.431.45–3.214.32–5.43
V6.54–9.5411.43–14.541.32–4.212.32–4.32
W10.43–20.3411.43–25.3436.32–45.6533.43–50.32
X23.1124.1748.3250.65

Average5.80–9.046.68–10.0412.95–17.3916.82–21.48