Files
hex/android/benchmark/AndroidManifest.xml
Ryan Vogel 23baa80aaa Measure on-device Android dictation across four languages
Replace the phase-one single-phrase evidence with a fixed 40-clip multilingual, partly music-mixed corpus. Record per-clip errors, load, warm inference, memory, two-core sensitivity and delivered platform-service results separately. Keep an independent foreground editor and media track for mic-to-insertion verification.

Verified on an API 35 ARM64 Studio emulator: Whisper tiny/base/small, Parakeet TDT v3 int8 and strictly on-device Android recognition. Small in its short window made 15/285 word errors; the platform service returned 33/40. Both native runners also decoded the notebook clip after tightening the PCM input checks.
2026-11-20 12:26:12 -05:00

9 lines
592 B
XML

<manifest xmlns:android="http://schemas.android.com/apk/res/android" package="local.hex.measure">
<uses-sdk android:minSdkVersion="34" android:targetSdkVersion="35" />
<uses-permission android:name="android.permission.RECORD_AUDIO" />
<queries><intent><action android:name="android.speech.RecognitionService" /></intent></queries>
<application android:theme="@android:style/Theme.Material.Light.NoActionBar" android:label="Audio benchmark">
<activity android:name=".SystemProbe" android:launchMode="singleTop" android:exported="true" />
</application>
</manifest>