Машинное обучение активно внедряется в разработки компаний уровня Google [1],
Facebook [2], Netflix [3] и др. Основополагающими причинами взрывного роста …

… audio-visual ASR. As an application of speaker adaptation, we consider the problem of
automatic recognition … of audio-visual ASR, and on what we view as open problems in this area. …

… ; and b The design of audio-visual information fusion algorithms … audio-only LVCSR systems,
under all possible audio-visual … art in audio-visual ASR by seriously tackling the problem of …

… audio-visual models the audio signal dominates, because speech recognition is a significantly
easier problem … babble noise with 0 dB SNR to the audio stream with probability pn ¼ 0:…

… audio pattern recognition tasks. Previous researchers have previously investigated transfer
learning for audio … To solve this problem, we design a balanced sampling strategy to train …

… We showed in [5] that the use of all features for classification does not always produce
good performance for the audio classification problems of our interest. This in turn leads to the …

… In Section III, we tackle the problem of integrating the information obtained from the …
audio-visual speech recognition systems. Finally, results on a multispeaker digit strings recognition …

… Next, we examine the available approaches to solving the problem of machine understanding
of human affective behavior occurring in real-world settings. We finally outline some …

… Summary and Conclusion This paper dealt with the problem of audio-visual speech integration
given that the relative reliabilities of the two modalities may fluctuate due to corruption in …

… audio time series with a short-time Fourier transform to create a spectrogram which was used
as an input to the CNN. The problem … approaches to the audio classification problem. The …