Abstract | ||
---|---|---|
The latency of current mobile devicesu0027 touchscreens is around 100ms and has widely been explored. Latency down to 2ms is noticeable, and latency as low as 25ms reduces usersu0027 performance. Previous work reduced touch latency by extrapolating a fingeru0027s movement using an ensemble of shallow neural networks and showed that predicting 33ms into the future increases usersu0027 performance. Unfortunately, this prediction has a high error. Predicting beyond 33ms did not increase participantsu0027 performance, and the error affected the subjective assessment. We use more recent machine learning techniques to reduce the prediction error. We train LSTM networks and multilayer perceptrons using a large data set and regularization. We show that linear extrapolation causes an 116.7% higher error and the previously proposed ensembles of shallow networks cause a 26.7% higher error compared to the LSTM networks. The trained models, the data used for testing, and the source code is available on GitHub. |
Year | Venue | Field |
---|---|---|
2017 | MobileHCI | Latency (engineering),Source code,Computer science,Touchscreen,Speech recognition,Software,Multilayer perceptron,Artificial intelligence,Artificial neural network,Lag,Perceptron,Machine learning |
DocType | Citations | PageRank |
Conference | 4 | 0.43 |
References | Authors | |
15 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Niels Henze | 1 | 1262 | 108.47 |
Sven Mayer | 2 | 188 | 27.30 |
Huy Viet Le | 3 | 95 | 13.02 |
Valentin Schwind | 4 | 143 | 21.22 |