GRACED: A Novel Fragile Watermarking for Speech Based on Endpoint Detection

Table 1

SNR (dB) between original speech and watermarked speech.

SNR (dB)	Ours	Ref. [14]	Ref. [16]	Ref. [25]
Male	57.43	37.87	29.57	30.37
Female	61.41	33.71	28.05	35.68
Child	64.61	30.51	27.55	39.02
Mean value	61.15	34.03	28.39	35.02