Speech-driven head motion generation from waveforms

Head motion generation task for speech-driven virtual agent animation is commonly explored with handcrafted audio features, such as MFCCs as input features, plus additional features, such as energy and F0 in the literature. In this paper, we study the direct use of speech waveform to generate head motion. We claim that creating a task-specific feature from waveform to generate head motion leads to better performance than using standard acoustic features to generate head motion overall. At the same time, we completely abandon the handcrafted feature extraction process, leading to more effectiveness. However, the difficulty of creating a task-specific feature from waveform is their staggering quantity of irrelevant information, implicating potential cumbrance for neural network training. Thus, we apply a canonical-correlation-constrained autoencoder (CCCAE), where we are able to compress the high-dimensional waveform into a low-dimensional embedded feature, with the minimal error in reconstruction, and sustain the relevant information with the maximal cannonical correlation to head motion. We extend our previous research by including more speakers in our dataset and also adapt with a recurrent neural network, to show the feasibility of our proposed feature. Through comparisons between different acoustic features, our proposed feature, WavCCCAE, shows at least a 20% improvement in the correlation from the waveform, and outperforms the popular acoustic feature, MFCC, by at least 5% respectively for all speakers. Through the comparison in the feedforward neural network regression (FNN-regression) system, the WavCCCAE-based system shows comparable performance in objective evaluation. In long short-term memory (LSTM) experiments, LSTM-models improve the overall performance in normalised mean square error (NMSE) and CCA metrics, and adapt the WavCCCAEfeature better, which makes the proposed LSTM-regression system outperform the MFCC-based system. We also re-design the subjective evaluation, and the subjective results show the animations generated by models where WavCCCAEwas chosen to be better than the other models by the participants of MUSHRA test.