Digital Signal Processing Reference
In-Depth Information
Figure 2-5. SEGSNR and WER Results for Reference Channel 3 Microphone (Chan3) and
Array Processing/Beamforming Scenarios using 67 speakers from the CU-Move Corpus. Bar
graph represents SegSNR in dB (using the left side scale), and line plot represents Avg. word
error rate improvement (in %) (using the right side scale).
Search WWH ::




Custom Search