Forward Attention in Sequence-to-sequence Acoustic Modelling for Speech Synthesis
This addresses alignment instability in speech synthesis systems, which is an incremental improvement over existing attention mechanisms.
The paper tackles the problem of unstable alignment in sequence-to-sequence speech synthesis by proposing a forward attention method that enforces monotonic alignment between phone and acoustic sequences, achieving faster convergence and higher stability than baseline attention methods while also improving speech naturalness and speed control.
This paper proposes a forward attention method for the sequenceto- sequence acoustic modeling of speech synthesis. This method is motivated by the nature of the monotonic alignment from phone sequences to acoustic sequences. Only the alignment paths that satisfy the monotonic condition are taken into consideration at each decoder timestep. The modified attention probabilities at each timestep are computed recursively using a forward algorithm. A transition agent for forward attention is further proposed, which helps the attention mechanism to make decisions whether to move forward or stay at each decoder timestep. Experimental results show that the proposed forward attention method achieves faster convergence speed and higher stability than the baseline attention method. Besides, the method of forward attention with transition agent can also help improve the naturalness of synthetic speech and control the speed of synthetic speech effectively.