Voice Recognition Apps: The Best in Assistive Tech
In a Staunton Media Lab (SML) broadcast streamed live on April 27, Audio Director Coley Evans, President Steve O’Keefe, and special guest Max Cross discussed (and demonstrated on a Chromebook) their preferred speech-to-text and screen capture software, with emphasis on their usability by the visually and hearing-impaired. We’ve covered the screen capture part of the broadcast in our previous blog post, and this one will focus on speech-to-text/transcription/voice recognition apps.
Assistive tech tools need improvement
Advancements in artificial Intelligence (AI), and voice- and image-recognition technology are making the world more inclusive for the visually impaired, notes Andrew Williams writing for Alphr, “helping them to interact with their surroundings.” True, but assistive tech still has ways to go, and voice recognition apps are no exception.
In his recent article for the UK-based TechWorld, Terry Hawkins, head of B2B solutions at the Royal National Institute of Blind People (RNIB), urged businesses to follow suit of the likes of giants like Facebook and Twitter to make online tech more accessible, “to make them better for all users,” and the blind and partially sighted in particular.
Working with blind and partially sighted people on a daily basis, we know all too well the struggle many face when dealing with websites and apps which aren’t accessible. Text-to-speech tools (which convert text into spoken words) and braille displays help this community better understand what is being shown on sites but, unfortunately, not all online services are well suited to the tools these people are using. And this needs to change.
Hawkins suggests improvements such as clear labeling of all elements, including buttons, for screen readers, uncluttered layouts, and making apps more mobile-friendly.
During the SML broadcast, Evans demonstrated his favorite voice recognition program developed in Russia, SpeechPad PW, which is open-source and runs on any device in the Chrome browser in Windows, Mac, and Linux OS. Evans said he likes features like typing format choice of all capital letters, or capital and lowercase; how easy it is to navigate, including using the Record feature; the reliability of recording (it won’t shut off after a few sentences like some other programs); and the accuracy of transcription, including punctuation.
According to Evans, with SpeechPad PW, “the results are superior” to those with other options, like Dragon Naturally Speaking. “This program is a lot better than you might think at first,” O’Keefe agreed during the demo, noting that one of the program’s “terrific uses” would be having a conversation with someone who is hearing impaired. Another would be to get an editable draft of, say, a homework assignment by a visually impaired person.
Here are some of the features:
- The Undo button removes the last entered sentence from the output field
- The punctuation buttons allow you to insert punctuation into the text
- If Execute voice commands checkbox is checked, then you can use voice commands
- The following words are automatically replaced: “period” (or “full stop”), “comma,” “question mark” and “exclamation mark”
- If the Transfer to clipboard checkbox is checked, the spoken sentences will go directly to the system clipboard and not the output field. (This checkbox is enabled if the SpeechPad Chrome extension is installed.)
- To change speech-recognition language, choose one from the drop-down menu, or register and add the desired speech input language in the User account.
Transcribing audio files
- The Transcription button shows or hides the audio recognition panel
- SpeechPad can recognize speech embedded in HTML5 video and audio
- For YouTube clips, specify the YouTube record ID before you record
Web and Windows OS use
- For use on websites, installing the SpeechPad Chrome extension
- SpeechPad Windows integration module works in any Windows application
- The Translation button activates a panel for translating speech input into other languages
- Check the simultaneous translation checkbox on that panel to see the translation of the spoken sentences into the selected language as you record
- Click on the extension icon and select a language in the drop-down list
Alternatives to SpeechPad
Below are three alternatives to SpeechPad PW, all also run in Chrome browser and are free:
A word processor with built-in multi-language speech recognition feature. Download from Chrome Web Store.
A text editor app, described as a “dictation notepad.” Download from Chrome Web Store.
- SpeechTexter‘s built-in speech recognition system promises accuracy with low noise level and good mike quality, as well as support for 30 languages. Download from Chrome Web Store.
Additional voice-recognition apps
Here’s a shortlist of some more popular free voice-recognition apps for Apple and Android platforms:
- Speech To Text (Xenom Apps). Speech-to-text for Android. Text can be sent in a text message or email, or copied and pasted into another app.
- Dragon Mobile Assistant (Nuance Communications, Inc.) An Android “virtual assistant” voice-recognition app that allows you to send and receive text messages and emails, post social media updates, maintain a calendar, search the web, and more. You can use it even if your screen is locked; create your voiceprint so that the app only works with your voice; and choose and voice and a name for your “assistant.”
- Dragon Dictation (Nuance Communications, Inc.) Apple app powered by Dragon Naturally Speaking that allows you to instantly see your text or email messages, and more.
- Voice Texting Pro (Sparkling Apps). For Apple, with features similar to Speech To Text’s. You can add languages for $.99.
- Evernote uses the built-in Android speech recognition software. Saves audio and text file together for easy tracking; has an in-app chat.
Finally, you can find more app suggestions and reviews on the U.K.’s Hearing Link site, Hollie Wicks’ list “Useful apps for hearing loss,” published last year. The “List of Computer Screen Readers for Visually Impaired” by Disabled World is another comprehensive resource we suggest.
Below is the April 27 SML broadcast:
Image source: nevarpp/123RF Stock Photo