Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 38 additions & 4 deletions plugins/Speech-to-Text/src/main/assets/docs/index.html
Original file line number Diff line number Diff line change
Expand Up @@ -51,10 +51,22 @@ <h2>Using it</h2>
model; otherwise the raw transcript is inserted.</li>
</ul>

<h2>On-device vs. cloud recognition</h2>
<p>The plugin prefers Android's <b>on-device</b> speech recognition. If the
device lacks an offline recognizer, Android may fall back to its network
recognition service.</p>
<h2>Languages and offline packs</h2>
<p>The plugin prefers Android's <b>on-device</b> speech recognition and asks
for the language Code on the Go itself is running in, so switching the app's
language switches the language you dictate in.</p>
<p>If the offline pack for that language is missing, the plugin asks the
recognizer what it does have and recovers without ending the capture:</p>
<ul>
<li>a pack for <b>another region</b> of the same language is used right away
(an es-US pack understands es-ES speech);</li>
<li>otherwise the missing pack is <b>requested for next time</b> and this
capture continues <b>online</b>, so the words you just spoke still land.
Pack downloads are best effort — Android does not always report them — so
nothing here waits on one.</li>
</ul>
<p>Either way you get one message and one prompt to speak again, and the
microphone button ignores taps while a capture is still running.</p>

<h2>Privacy</h2>
<div class="note">
Expand All @@ -73,6 +85,28 @@ <h2>Troubleshooting</h2>
recognizer; install/enable Google's speech services.</li>
<li><b>Permission prompt didn't appear</b> — grant Microphone under
Settings → Apps → CodeOnTheGo → Permissions, then tap again.</li>
<li><b>"No offline pack for …, so … is being used instead"</b> — a pack for
another region of the same language was used. Add the exact language under
Settings → System → Languages &amp; input → On-device recognition if you
want it.</li>
<li><b>"No offline pack for … yet — fetching it for next time"</b> — speak
again; this capture is recognized online. If the pack arrives, later
captures are offline again. To be sure of it, install the language under
Settings → System → Languages &amp; input → On-device recognition.</li>
<li><b>"Voice input couldn't use … on this device"</b> — no offline pack, so
the plugin is retrying over the network. Just speak again.</li>
<li><b>"Voice input can't use … on this device"</b> — offline and online both
failed for that language. Install the pack from that same screen, or switch
the app's language.</li>
<li><b>"Speech recognition is still finishing the previous request"</b> — a
capture is already running; wait for it to finish, then tap again.</li>
<li><b>"Speech recognition doesn't support …"</b> — the recognizer has no
model for that language at all, offline or online. Switch your device
language to a supported one.</li>
<li><b>"Voice input stopped unexpectedly"</b> — the device's speech service
failed in a way the plugin can't recover from. Tap the microphone to start
over; if it keeps happening, check for an update to the speech service
(Speech Services by Google, or your device's own).</li>
</ul>
</body>
</html>
Loading