Leveraging GANs to Improve Continuous Path Keyboard Input Models
This addresses the problem of data scarcity for developers of continuous path keyboard systems, though it is incremental as it applies an existing method (GANs) to a specific domain.
The paper tackled the challenge of training robust continuous path keyboard input models by using GANs to generate synthetic data, resulting in a substantial boost in accuracy at a 5:1 GAN-to-real ratio.
Continuous path keyboard input has higher inherent ambiguity than standard tapping, because the path trace may exhibit not only local overshoots/undershoots (as in tapping) but also, depending on the user, substantial mid-path excursions. Deploying a robust solution thus requires a large amount of high-quality training data, which is difficult to collect/annotate. In this work, we address this challenge by using GANs to augment our training corpus with user-realistic synthetic data. Experiments show that, even though GAN-generated data does not capture all the characteristics of real user data, it still provides a substantial boost in accuracy at a 5:1 GAN-to-real ratio. GANs therefore inject more robustness in the model through greatly increased word coverage and path diversity.