Faster and Less Error-Prone: Supplementing an Accessible Keyboard with Speech Input

Abstract

Swarachakra is an Abugida text input keyboard available in 12 Indian languages. We enhanced an accessible version of Swarachakra Marathi with speech input. However, speech input could be error-prone, and especially so for languages where speech recognition technologies are new. Such errors could either slow the user down due to the need for editing, or go unnoticed, leading to high uncorrected error rates. We therefore conducted a withinsubject empirical study to compare the user performance of keyboard-only input method with keyboard+speech input method with 11 novice visually impaired users. We found that keyboard+speech input was almost 11 times faster, reaching 182 characters per minute, and had a lower uncorrected error rate than the keyboard-only input, and in spite of having higher corrected error rates. Though we used a wide variety of phrases in our study, we observed that all phrases were faster on average with the keyboard+speech input method. To the best of our knowledge, ours is the first empirical study to evaluate the performance of speech enabled text input in Marathi for visually impaired people. This is the highest reported speed by visually impaired users in any Indian language.

A Framework to Improve Web Form Accessibility for the Visually Impaired

“…As our test sample consisted of Arabic speakers, the possibility of errors may increase due to the relatively immature state of speech recognition technologies for the Arabic language [32]. Consistent with another study [33], which found that correcting error rates using both the keyboard and speech were higher compared to the keyboard-only condition, we allowed participants to use both methods to enter data into the web form. Regarding input time for both methods, it is natural for speech input to be faster than writing using the keyboard.…”

“…Regarding input time for both methods, it is natural for speech input to be faster than writing using the keyboard. This was demonstrated by the only person who chose speech input, and it was supported by a study [33], stating that speech input was nearly five times faster than keyboard input. However, the time required to fix the entered text may be twice as long.…”

edATLAS: An Efficient Disambiguation Algorithm for Texting in Languages with Abugida Scripts

“…Big data analytics of 3 Samsung Galaxy smartphones, all launched in 2019-2020, reveal that across top-5 Hindi speaking states of India (by native speaker population [3]), the average number of Hindi language users who prefer to type in native layouts exceeds the number of users in the same demography who prefer transliteration layouts [4] by a very large margin, as seen in Figure 2. Even with such a significant user base observed for other abugida script languages, there are multiple pain points for an end-user to text in native scripts [5]. Firstly, singular layouts with regional scripts are not pervasive as it is not feasible to accommodate 50+ native characters in any single keyboard layout.…”