Skip to main content
Figure 17 | EURASIP Journal on Advances in Signal Processing

Figure 17

From: Multi-pose lipreading and audio-visual speech recognition

Figure 17

Performance of different AV-ASR systems with a corrupted audio stream with 7 dB of SNR. Mean word accuracy for audio and audio-visual systems with different visual streams and classifiers. The audio stream is a corrupted with 7 dB of SNR, while different visual streams are considered: the ideal frontal (F) views of the speaker, the original lateral (L) views at 30°, 60° and 90° of head rotation and the corresponding streams after pose normalization.

Back to article page