Self-Content-Based Audio Inpainting:
*******************************

The following is a legend, relating each file name (a number) to its configuration parameters:

Story scenario:
S1.wav	The audio sequence, WITHOUT audio data gaps.
S2.wav	The pierced audio sequence. Audio gaps appear as silence.
S3.wav	The sequence inpainted using the method by Adler et al. [5] (Did not appear in the subjective survey, due to its extremely poor performance)
S4.wav	The sequence inpainted using the G.723.1 standard.
S5.wav	The sequence inpainted using the G.711 standard.
S6.wav	Our basic audio inpainting method.
S7.wav	Our audio inpating method (as in S6), with incorporation of prior knowledge term R.
S8.wav	Our audio inpating method (as in S6), with incorporation of prior knowledge term R and pitch modification.
S9.wav	The full original audio sequence, including the part which was later pierced (identical to S1.wav) and the rest, which was used for extracting example ABs.

Conversation scenario:
C1.wav	The conversation audio sequence, WITHOUT audio data gaps.
C2.wav	The pierced audio sequence. Audio gaps appear as silence.
C3.wav	The sequence inpainted using the method by Adler et al. [5]
C4.wav	The sequence inpainted using the G.723.1 standard.
C5.wav	The sequence inpainted using the G.711 standard.
C6.wav	Our basic audio inpainting method.
C7.wav	Our audio inpating method (as in S6), with incorporation of prior knowledge term R.
C8.wav	The full original conversation audio sequence, including the last 2 minutes that was later pierced (identical to C1.wav) and the rest, which was only used for extracting example ABs.