Android/smart_human
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

临时提交

Browse Source
This commit is contained in:
林若思 2025-12-30 15:49:56 +08:00
parent 738b39e9a0
commit 885dcbd044
20 changed files with 439 additions and 233 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

0
app/CMakeLists.txt Normal file
View File

22
app/build.gradle
View File

@ -16,6 +16,13 @@ android {
versionName "1.0.0"
multiDexEnabled true
testInstrumentationRunner "androidx.test.runner.AndroidJUnitRunner"
externalNativeBuild {
cmake {
cppFlags ""
}
}
ndk {
// 设置支持的SO库架构，第三方给的so库哪几种架构，就配置这几种架构
abiFilters 'armeabi-v7a', 'arm64-v8a'
@ -76,7 +83,7 @@ android {
enable = true
}
ndkVersion '29.0.13846066 rc3'
ndkVersion '27.0.12077973'
applicationVariants.configureEach { variant ->
if (variant.buildType.name != 'release') return
@ -88,6 +95,18 @@ android {
}
}
sourceSets {
main {
jniLibs.srcDirs = ['src/main/libs']
}
}
externalNativeBuild {
cmake {
path "CMakeLists.txt"
}
}
}
protobuf {
@ -162,4 +181,5 @@ dependencies {
implementation libs.androidautosize
implementation files('libs/sherpa19.aar')
}

12
app/proguard-rules.pro vendored
View File

@ -278,3 +278,15 @@
<fields>;
<methods>;
}
-keepclassmembers class android.media.AudioManager {
int maxActivePaths;
}
-keep class com.k2fsa.sherpa.onnx.** { *; }
-keepclassmembers class com.k2fsa.sherpa.onnx.** {
native <methods>;
<init>(...);
}
-keep interface com.k2fsa.sherpa.onnx.** { *; }

6
app/src/main/AndroidManifest.xml
View File

@ -52,7 +52,7 @@
tools:targetApi="31">
<activity
android:name="com.zs.smarthuman.ui.SplashActivity"
android:name=".ui.MainActivity"
android:exported="true"
android:theme="@style/Theme.Splash"
android:screenOrientation="portrait">
@ -66,10 +66,10 @@
</intent-filter>
</activity>
<activity
<!-- <activity
android:name="com.zs.smarthuman.ui.MainActivity"
android:screenOrientation="portrait"/>
-->
<activity
android:name="com.zs.smarthuman.ui.ActivateActivity"
android:screenOrientation="portrait"/>

BIN
app/src/main/assets/ten-vad.onnx Normal file
View File

Binary file not shown.

2
app/src/main/java/com/zs/smarthuman/http/ApiService.kt
View File

@ -14,6 +14,6 @@ class ApiService {
const val GET_USER_INFO_URL = "iot/info/getUserInfo"
const val UPLOAD_RECORD_VOICE_URL = "/iot/chat/test"
const val UPLOAD_RECORD_VOICE_URL = "/iot/chat"
}
}

8
app/src/main/java/com/zs/smarthuman/im/chat/MessageContentType.java
View File

@ -16,7 +16,13 @@ public enum MessageContentType {
/**
* 接收后台音频响应指令
*/
RECEIVE_VOICE_STREAM(2);
RECEIVE_VOICE_STREAM(2),
/**
* 更新用户信息
*/
UPDATE_INFO(3);
private int msgContentType;

274
app/src/main/java/com/zs/smarthuman/sherpa/VoiceController.kt
View File

@ -2,179 +2,190 @@ package com.zs.smarthuman.sherpa
import android.content.res.AssetManager
import android.util.Log
import kotlinx.coroutines.*
import kotlin.math.sqrt
class VoiceController(
assetManager: AssetManager,
private val onWakeup: () -> Unit,
private val onFinalAudio: (FloatArray) -> Unit,
private val idleTimeoutSeconds: Int = 15,
private val maxRecordingSeconds: Int = 10, // ✅ 最大录音时长
private val maxRecordingSeconds: Int = 10,
private val onStateChanged: ((VoiceState) -> Unit)? = null,
private val stopBackendAudio: (() -> Unit)? = null
) {
private val TAG = "VoiceController"
private val sampleRate = 16000
/* ================= 状态 ================= */
private var state: VoiceState = VoiceState.WAIT_WAKEUP
set(value) {
field = value
Log.d(TAG, "➡ State = $value")
onStateChanged?.invoke(value)
Log.d(TAG, "当前状态: $value")
}
var isPlaying = false
private set
/* ================= 唤醒 ================= */
private val WAKEUP_DISCARD_MS = 600L
private val WAKEUP_COOLDOWN_MS = 1500L
private var wakeupDiscardUntil = 0L
private var lastWakeupTime = 0L
private val wakeupManager = WakeupManager(assetManager) {
val now = System.currentTimeMillis()
if (now - lastWakeupTime < WAKEUP_COOLDOWN_MS) {
Log.d(TAG, "⚠️ 唤醒过于频繁，忽略")
return@WakeupManager
}
lastWakeupTime = now
Log.d(TAG, "🔥 唤醒触发")
Log.d(TAG, "🔥 WakeWord detected")
stopBackendAudio?.invoke()
isPlaying = false
audioBuffer.clear()
preBuffer.clear()
vadManager.reset()
vadStarted = false
vadEndPending = false
wakeupDiscardUntil = now + WAKEUP_DISCARD_MS
resetAll()
state = VoiceState.PLAYING_PROMPT
onWakeup()
playLocalPrompt()
}
/* ================= VAD ================= */
/* ================= VAD（只负责 START） ================= */
private val vadManager = VadManager(
assetManager,
onSpeechStart = { onVadSpeechStart() },
onSpeechEnd = { onVadSpeechEnd() }
onSpeechStart = { onVadStart() },
onSpeechEnd = { /* ❌ 不再用于结束 */ }
)
/* ================= 音频缓存 ================= */
private val audioBuffer = mutableListOf<Float>()
private val preBuffer = ArrayDeque<Float>()
private val PRE_BUFFER_SIZE = 16000
private val PRE_BUFFER_SIZE = sampleRate / 2 // 500ms
/* ================= 时间 ================= */
private var idleTimer = 0L
private var recordingStartTime = 0L
private var vadStarted = false
private var vadEndPending = false
private var vadEndTime = 0L
private val END_SILENCE_MS = 1000L
/* ================= 外部音频输入 ================= */
private var recordingStartTime = 0L // ✅ 记录录音开始时间
/* ================= RMS 结束判定 ================= */
private var silenceStartMs = 0L
private val SILENCE_END_MS = 1200L // 静音多久算一句结束
private val RMS_SILENCE_THRESHOLD = 0.01f // 静音能量阈值
private val MIN_SPEECH_DURATION_MS = 800L // 最短有效语音
/* ================= 音频入口 ================= */
fun acceptAudio(samples: FloatArray) {
cachePreBuffer(samples)
wakeupManager.acceptAudio(samples)
val now = System.currentTimeMillis()
if (now < wakeupDiscardUntil) return
if (state == VoiceState.UPLOADING ||
state == VoiceState.PLAYING_PROMPT ||
state == VoiceState.PLAYING_BACKEND
) return
when (state) {
VoiceState.WAIT_SPEECH -> {
if (state == VoiceState.WAIT_SPEECH) {
cachePreBuffer(samples)
vadManager.accept(samples)
return
}
VoiceState.RECORDING -> {
if (state != VoiceState.RECORDING) return
// ===== RECORDING =====
audioBuffer.addAll(samples.asList())
vadManager.accept(samples)
idleTimer = now
// ✅ 最大录音时长判断
val now = System.currentTimeMillis()
// 1️⃣ 最大录音兜底
if (now - recordingStartTime >= maxRecordingSeconds * 1000) {
Log.d(TAG, "⚠️ 达到最大录音时长，自动结束录音")
Log.w(TAG, "⏱ Max recording reached")
finishSentence()
return
}
if (vadEndPending && now - vadEndTime >= END_SILENCE_MS) {
// 2️⃣ RMS 静音结束（核心）
val rms = calcRms(samples)
// Log.d(TAG, "RMS_DEBUG", "rms=${"%.4f".format(rms)}")
if (rms < RMS_SILENCE_THRESHOLD) {
if (silenceStartMs == 0L) {
silenceStartMs = now
} else if (now - silenceStartMs >= SILENCE_END_MS) {
Log.d(TAG, "🔇 RMS silence end")
finishSentence()
}
}
else -> Unit
} else {
silenceStartMs = 0L
}
}
/* ================= 提示音 ================= */
/* ================= VAD START ================= */
private fun onVadStart() {
if (state != VoiceState.WAIT_SPEECH) return
private val PROMPT_DURATION_MS = 3000L
private var promptFallbackJob: Job? = null
Log.d(TAG, "🎤 VAD START")
vadStarted = true
state = VoiceState.RECORDING
recordingStartTime = System.currentTimeMillis()
silenceStartMs = 0L
audioBuffer.addAll(preBuffer)
preBuffer.clear()
}
/* ================= 结束录音 ================= */
private fun finishSentence() {
val speakTime = System.currentTimeMillis() - recordingStartTime
if (!vadStarted || speakTime < MIN_SPEECH_DURATION_MS) {
Log.d(TAG, "⛔ Speech too short, ignore")
resetToWaitSpeech()
return
}
val finalAudio = audioBuffer.toFloatArray()
audioBuffer.clear()
state = VoiceState.UPLOADING
Log.d(TAG, "⬆ Upload audio len=${finalAudio.size}")
onFinalAudio(finalAudio)
}
/* ================= 播放回调 ================= */
fun onPlayStartPrompt() {
if (state == VoiceState.PLAYING_PROMPT) return
isPlaying = true
state = VoiceState.PLAYING_PROMPT
promptFallbackJob?.cancel()
promptFallbackJob = CoroutineScope(Dispatchers.Main).launch {
delay(PROMPT_DURATION_MS)
if (state == VoiceState.PLAYING_PROMPT) {
Log.w(TAG, "⚠️ 提示音 complete 丢失，fallback")
onPlayEndPrompt()
}
}
}
fun onPlayEndPrompt() {
promptFallbackJob?.cancel()
if (state != VoiceState.PLAYING_PROMPT) return
isPlaying = false
state = VoiceState.WAIT_SPEECH
idleTimer = System.currentTimeMillis()
Log.d(TAG, "提示音结束 → WAIT_SPEECH")
}
/* ================= Backend ================= */
fun onPlayStartBackend() {
isPlaying = true
state = VoiceState.PLAYING_BACKEND
}
fun onPlayEndBackend() {
if (state != VoiceState.PLAYING_BACKEND) return
isPlaying = false
state = VoiceState.WAIT_WAKEUP
state = VoiceState.WAIT_SPEECH
idleTimer = System.currentTimeMillis()
}
/* ================= 上传回调 ================= */
fun onUploadFinished(success: Boolean) {
if (state != VoiceState.UPLOADING) return
state = if (success) VoiceState.PLAYING_BACKEND else VoiceState.WAIT_SPEECH
}
/* ================= Idle ================= */
fun checkIdleTimeout() {
if (state != VoiceState.WAIT_SPEECH) return
if (System.currentTimeMillis() - idleTimer > idleTimeoutSeconds * 1000) {
reset()
resetAll()
}
}
fun reset() {
state = VoiceState.WAIT_WAKEUP
/* ================= Reset ================= */
private fun resetToWaitSpeech() {
audioBuffer.clear()
preBuffer.clear()
vadManager.reset()
vadStarted = false
vadEndPending = false
wakeupDiscardUntil = 0L
recordingStartTime = 0L // ✅ 重置录音开始时间
Log.d(TAG, "reset → WAIT_WAKEUP")
silenceStartMs = 0L
state = VoiceState.WAIT_SPEECH
idleTimer = System.currentTimeMillis()
}
private fun resetAll() {
audioBuffer.clear()
preBuffer.clear()
vadManager.reset()
vadStarted = false
silenceStartMs = 0L
state = VoiceState.WAIT_WAKEUP
}
fun release() {
@ -182,84 +193,7 @@ class VoiceController(
wakeupManager.release()
}
/* ================= VAD 回调 ================= */
private fun onVadSpeechStart() {
if (state != VoiceState.WAIT_SPEECH) return
vadStarted = true
state = VoiceState.RECORDING
audioBuffer.addAll(preBuffer)
idleTimer = System.currentTimeMillis()
recordingStartTime = System.currentTimeMillis() // ✅ 记录录音开始时间
Log.d(TAG, "VAD开始 → RECORDING")
}
private fun onVadSpeechEnd() {
if (state != VoiceState.RECORDING) return
vadEndPending = true
vadEndTime = System.currentTimeMillis()
}
/* ================= 录音结束 & 判定 ================= */
private fun finishSentence() {
vadEndPending = false
state = VoiceState.WAIT_WAKEUP
val finalAudio = audioBuffer.toFloatArray()
audioBuffer.clear()
if (isValidUserSpeech(finalAudio)) {
onFinalAudio(finalAudio)
Log.d(TAG, "✅ 录音有效，上传")
} else {
Log.d(TAG, "❌ 噪音/旁人语音，丢弃")
}
}
/* ================= 关键判定函数 ================= */
private fun isValidUserSpeech(audio: FloatArray): Boolean {
if (!vadStarted) {
Log.d(TAG, "❌ VAD 未触发")
return false
}
// 1️⃣ 时长：>= 600ms（非常宽松）
val durationMs = audio.size * 1000f / 16000f
if (durationMs < 600f) {
Log.d(TAG, "❌ 太短: ${durationMs}ms")
return false
}
// 2️⃣ 计算 RMS（真实设备极低）
var sum = 0f
var peak = 0f
for (v in audio) {
val a = kotlin.math.abs(v)
sum += a * a
if (a > peak) peak = a
}
val rms = kotlin.math.sqrt(sum / audio.size)
Log.d(TAG, "🎤 RMS=$rms peak=$peak duration=${durationMs}ms")
// 3️⃣ 只排除“纯底噪”
// 实测：环境底噪 RMS 通常 < 0.001
if (rms < 0.002f && peak < 0.01f) {
Log.d(TAG, "❌ 纯环境噪声，丢弃")
return false
}
// ✅ 只要不是纯噪声，一律认为是人说话
Log.d(TAG, "✅ 判定为有效人声")
return true
}
/* ================= 工具 ================= */
private fun cachePreBuffer(samples: FloatArray) {
for (s in samples) {
preBuffer.addLast(s)
@ -269,7 +203,11 @@ class VoiceController(
}
}
private fun playLocalPrompt() {
onPlayStartPrompt()
private fun calcRms(audio: FloatArray): Float {
var sum = 0f
for (v in audio) {
sum += v * v
}
return sqrt(sum / audio.size)
}
}

1
app/src/main/java/com/zs/smarthuman/sherpa/VoiceState.kt
View File

@ -10,5 +10,6 @@ enum class VoiceState {
PLAYING_PROMPT, // 播放本地提示音
WAIT_SPEECH, // 等待用户说话
RECORDING, // 用户正在说话
UPLOADING, //音频上传中
PLAYING_BACKEND // 播放后台返回音频
}

15
app/src/main/java/com/zs/smarthuman/ui/ActivateActivity.kt
View File

@ -5,6 +5,7 @@ import android.graphics.drawable.AnimationDrawable
import android.graphics.drawable.Drawable
import android.view.View
import android.view.animation.LinearInterpolator
import androidx.lifecycle.lifecycleScope
import com.bumptech.glide.Glide
import com.bumptech.glide.request.target.ImageViewTarget
import com.bumptech.glide.request.transition.Transition
@ -14,8 +15,12 @@ import com.zs.smarthuman.bean.NetworkStatusEventMsg
import com.zs.smarthuman.bean.QRCodeResp
import com.zs.smarthuman.databinding.ActivityActivateBinding
import com.zs.smarthuman.http.ApiResult
import com.zs.smarthuman.im.chat.MessageContentType
import com.zs.smarthuman.im.chat.bean.SingleMessage
import com.zs.smarthuman.kt.startActivity
import com.zs.smarthuman.viewmodel.ActivateViewModel
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.launch
/**
* @description:二维码激活页面
@ -90,7 +95,15 @@ class ActivateActivity : BaseViewModelActivity<ActivityActivateBinding, Activate
}
override fun receivedIMMsg(msg: SingleMessage) {
when (msg.msgContentType) {
MessageContentType.ACTIVE_SUCCESS_ENTER_MAIN.msgContentType -> {
lifecycleScope.launch(Dispatchers.Main){
startActivity(MainActivity::class)
overridePendingTransition(0,0)
finish()
}
}
}
}
override fun onMessageEvent(event: Any) {

100
app/src/main/java/com/zs/smarthuman/ui/MainActivity.kt
View File

@ -9,6 +9,7 @@ import android.content.pm.PackageManager
import android.media.AudioFormat
import android.media.AudioRecord
import android.media.MediaRecorder
import android.media.audiofx.AcousticEchoCanceler
import android.os.Bundle
import android.os.Environment
import android.os.IBinder
@ -95,6 +96,7 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
PermissionUtils.permissionGroup(PermissionConstants.MICROPHONE)
.callback(object : PermissionUtils.FullCallback {
override fun onGranted(granted: List<String?>) {
initVoiceController()
initMicrophone()
startRecording()
}
@ -109,8 +111,6 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
}).request()
initVoiceController()
requestUserInfo()
initObserver()
}
@ -158,7 +158,7 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
voiceInfo = mutableListOf<VoiceBeanResp>().apply {
add(
VoiceBeanResp(
audioUrl = UserInfoManager.userInfo?.wakeUpAudioUrl ?: ""
audioUrl = /*UserInfoManager.userInfo?.wakeUpAudioUrl ?:*/ "https://static.seerteach.net/aidialogue/systemVoice/aliyun-nv.mp3"
)
)
}
@ -166,10 +166,11 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
},
onFinalAudio = { audio ->
Log.d("lrs", "检测到语音，长度=${audio.size}")
mViewModel?.uploadVoice(
AudioPcmUtil.pcm16ToBase64(AudioPcmUtil.floatToPcm16(audio)),
1
)
// mViewModel?.uploadVoice(
// AudioPcmUtil.pcm16ToBase64(AudioPcmUtil.floatToPcm16(audio)),
// 1
// )
loadLocalJsonAndPlay()
val file = File(
getExternalFilesDir(Environment.DIRECTORY_DOWNLOADS)!!.getAbsolutePath(),
"xxx.wav"
@ -192,6 +193,7 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
VoiceState.RECORDING -> Log.d("lrs", "当前状态: 正在录音")
VoiceState.PLAYING_PROMPT -> Log.d("lrs", "当前状态: 播放本地音频")
VoiceState.PLAYING_BACKEND -> Log.d("lrs", "当前状态: 播放后台音频")
else -> {}
}
},
@ -255,26 +257,38 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
if (audioRecord?.state != AudioRecord.STATE_INITIALIZED) {
Log.e("VoiceService", "Failed to initialize AudioRecord")
}
enableSystemAec(audioRecord!!)
}
private var aec: AcousticEchoCanceler? = null
private fun enableSystemAec(record: AudioRecord) {
if (!AcousticEchoCanceler.isAvailable()) {
Log.w("VoiceService", "System AEC not available")
return
}
aec = AcousticEchoCanceler.create(record.audioSessionId)
aec?.enabled = true
Log.d("VoiceService", "✅ System AEC enabled")
}
//开始录音
fun startRecording() {
isRecording = true
audioRecord?.startRecording()
lifecycleScope.launch(Dispatchers.IO) {
val buf = ShortArray(640)
val buf = ShortArray(512)
while (isRecording) {
val n = audioRecord?.read(buf, 0, buf.size) ?: 0
if (n > 0) {
val raw = FloatArray(n) { buf[it] / 32768f }
// 播放时 duck
val ducked = if (voiceController?.isPlaying == true) {
FloatArray(n) { raw[it] * 0.4f }
} else raw
voiceController?.acceptAudio(agc(ducked))
voiceController?.acceptAudio(raw)
}
voiceController?.checkIdleTimeout()
}
@ -291,23 +305,6 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
}
}
private fun agc(
input: FloatArray,
target: Float = 0.035f, // ⬅️ 降
maxGain: Float = 4f // ⬅️ 关键
): FloatArray {
var sum = 0f
for (v in input) sum += v * v
val rms = sqrt(sum / input.size)
if (rms < 1e-6) return input
val gain = (target / rms).coerceAtMost(maxGain)
return FloatArray(input.size) {
(input[it] * gain).coerceIn(-1f, 1f)
}
}
fun onUnityResourcesLoaded(message: String) { // 这是 Unity 调用的资源加载完成回调方法
lifecycleScope.launch(Dispatchers.Main) {
@ -315,7 +312,9 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
binding.flDigitalHuman.translationY = 0f
}
}
private var promptPlaying = false
private var backPlaying = false
fun onAudioProgressUpdated( // Unity 调用此方法传递音频进度
progress: Float,
state: Int,//0stop 2pause 1play 3complete 4loading 5error
@ -323,9 +322,9 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
word: String,
audioUrl: String
) {
val wakeupUrl = UserInfoManager.userInfo?.wakeUpAudioUrl ?: return
if (audioUrl != wakeupUrl) return
// val wakeupUrl = UserInfoManager.userInfo?.wakeUpAudioUrl ?: return
//
// if (audioUrl != wakeupUrl) return
when (state) {
1 -> { // play
@ -337,7 +336,6 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
3 -> { // complete
if (promptPlaying) {
Toaster.showShort("借宿了")
promptPlaying = false
voiceController?.onPlayEndPrompt()
}
@ -352,13 +350,43 @@ class MainActivity : BaseViewModelActivity<ActivityMainBinding, MainViewModel>()
text: String,
isFinal: Boolean
) {
if (state == 1) {
when (state) {
1 -> { // play
if (!backPlaying) {
backPlaying = true
voiceController?.onPlayStartBackend()
}
if (state == 3) {
}
3 -> { // complete
if (backPlaying) {
Toaster.showShort("借宿了")
backPlaying = false
voiceController?.onPlayEndBackend()
}
}
}
}
private fun loadLocalJsonAndPlay() {
lifecycleScope.launch(Dispatchers.IO) {
try {
val jsonStr = assets
.open("readEnd.json")
.bufferedReader()
.use { it.readText() }
UnityPlayerHolder
.getInstance()
.startTalking(jsonStr)
} catch (e: Exception) {
e.printStackTrace()
LogUtils.eTag("lrs", "loadLocalJsonAndPlay error: ${e.message}")
}
}
}
override fun onDestroy() {

181
app/src/main/java/com/zs/smarthuman/utils/PcmAudioWithAecManager.kt Normal file
View File

@ -0,0 +1,181 @@
package com.zs.smarthuman.utils
import android.annotation.SuppressLint
import android.media.AudioFormat
import android.media.AudioManager
import android.media.AudioRecord
import android.media.AudioTrack
import android.media.MediaRecorder
import android.os.Environment
import android.util.Base64
import android.util.Log
import com.blankj.utilcode.util.LogUtils
import com.zs.smarthuman.App
import com.zs.smarthuman.sherpa.VoiceController
import com.zs.smarthuman.utils.AudioDebugUtil
import java.io.File
import java.nio.ByteBuffer
import java.nio.ByteOrder
import java.util.concurrent.ArrayBlockingQueue
import kotlin.concurrent.thread
class PcmAudioWithAecManager(
private val voiceController: VoiceController
) {
companion object {
private const val TAG = "PcmAudioWithAecManager"
}
private val sampleRate = 16000
private val channelConfig = AudioFormat.CHANNEL_IN_MONO
private val audioFormat = AudioFormat.ENCODING_PCM_16BIT
private val bufferSize = AudioRecord.getMinBufferSize(sampleRate, channelConfig, audioFormat)
private var audioRecord: AudioRecord? = null
private var audioTrack: AudioTrack? = null
private val playQueue = ArrayBlockingQueue<ShortArray>(200)
@Volatile
var isRecording = false
@Volatile
private var isPlaying = false
private val rawBuffer = mutableListOf<Float>()
private val aecBuffer = mutableListOf<Float>()
init {
initAudioRecord()
initAudioTrack()
}
@SuppressLint("MissingPermission")
private fun initAudioRecord() {
audioRecord = AudioRecord(
MediaRecorder.AudioSource.VOICE_COMMUNICATION,
sampleRate,
channelConfig,
audioFormat,
bufferSize
)
}
private fun initAudioTrack() {
val trackBufferSize = AudioTrack.getMinBufferSize(
sampleRate,
AudioFormat.CHANNEL_OUT_MONO,
audioFormat
)
audioTrack = AudioTrack(
AudioManager.STREAM_MUSIC,
sampleRate,
AudioFormat.CHANNEL_OUT_MONO,
audioFormat,
trackBufferSize,
AudioTrack.MODE_STREAM
)
}
/** 开始录音并进行处理 */
fun startRecording() {
if (isRecording) return
isRecording = true
audioRecord?.startRecording()
thread {
val buffer = ShortArray(bufferSize / 2)
while (isRecording) {
val read = audioRecord?.read(buffer, 0, buffer.size) ?: 0
if (read > 0) {
// 录音数据处理
processCapture(buffer.copyOf(read))
}
voiceController.checkIdleTimeout()
}
}
}
/** 停止录音并保存 WAV */
fun stopRecording() {
isRecording = false
audioRecord?.stop()
// Log.d(TAG, "rawBuffer:${rawBuffer.size},${aecBuffer.size}")
// if (rawBuffer.isNotEmpty()) {
// val rawFile = File(
// App.getInstance().getExternalFilesDir(Environment.DIRECTORY_DOWNLOADS)!!.absolutePath,
// "raw.wav"
// )
// AudioDebugUtil.saveFloatPcmAsWav(rawBuffer.toFloatArray(), rawFile)
// Log.d(TAG, "RAW WAV saved: ${rawFile.absolutePath}")
// rawBuffer.clear()
// }
//
// if (aecBuffer.isNotEmpty()) {
// val aecFile = File(
// App.getInstance().getExternalFilesDir(Environment.DIRECTORY_DOWNLOADS)!!.absolutePath,
// "aec_clean.wav"
// )
// AudioDebugUtil.saveFloatPcmAsWav(aecBuffer.toFloatArray(), aecFile)
// Log.d(TAG, "AEC WAV saved: ${aecFile.absolutePath}")
// aecBuffer.clear()
// }
}
/** 开始播放队列音频 */
fun startPlaying() {
if (isPlaying) return
isPlaying = true
audioTrack?.play()
thread {
while (isPlaying) {
try {
val buffer = playQueue.take()
// 播放音频时不暂停录音，继续录音
audioTrack?.write(buffer, 0, buffer.size)
} catch (e: InterruptedException) {
e.printStackTrace()
}
}
}
}
/** 停止播放 */
fun stopPlaying() {
isPlaying = false
audioTrack?.stop()
}
/** 处理录音音频 */
private fun processCapture(buffer: ShortArray) {
// 保存原始 PCM
val floatRaw = FloatArray(buffer.size) { buffer[it] / 32768f }
rawBuffer.addAll(floatRaw.toList())
// 向 VoiceController 发送数据进行唤醒词检测
voiceController.acceptAudio(floatRaw)
}
/** 推入播放队列 */
private fun pushPlayBuffer(buffer: ShortArray) {
if (!playQueue.offer(buffer)) {
Log.w(TAG, "playQueue 满了，丢弃一帧")
}
}
/** 播放后台 Base64 音频 */
fun playBackendAudio(base64: String) {
val bytes = Base64.decode(base64, Base64.DEFAULT)
val shortBuffer = ShortArray(bytes.size / 2)
ByteBuffer.wrap(bytes).order(ByteOrder.LITTLE_ENDIAN).asShortBuffer().get(shortBuffer)
pushPlayBuffer(shortBuffer)
}
/** 释放资源 */
fun release() {
stopRecording()
stopPlaying()
audioRecord?.release()
audioTrack?.release()
}
}

16
app/src/main/java/com/zs/smarthuman/utils/SerialNumberUtil.kt
View File

@ -25,14 +25,14 @@ object SerialNumberUtil {
* 外部调用，获取最终序列号
*/
fun getSerialNumber(): String {
// for (key in snKeys) {
// val sn = getProp(key)
// if (!sn.isNullOrBlank()) {
// return limitSerialDigit(sn)
// }
// }
// return ""
return "zd09312051870556"
for (key in snKeys) {
val sn = getProp(key)
if (!sn.isNullOrBlank()) {
return limitSerialDigit(sn)
}
}
return ""
// return "zd09312051870556"
}
/**

BIN
app/src/main/res/drawable/icon_people_speaking.webp Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 1.9 MiB

7
app/src/main/res/layout/activity_main.xml
View File

@ -31,5 +31,12 @@
app:layout_constraintTop_toTopOf="parent"/>
<com.shuyu.gsyvideoplayer.video.StandardGSYVideoPlayer
android:id="@+id/standardGSYVideoPlayer"
android:layout_width="0dp"
android:layout_height="0dp"
app:layout_constraintBottom_toBottomOf="parent"
app:layout_constraintEnd_toEndOf="parent"/>
</androidx.constraintlayout.widget.ConstraintLayout>

2
unityLibrary/build.gradle
View File

@ -8,7 +8,7 @@ dependencies {
android {
namespace "com.unity3d.player"
ndkPath "D:\\Android\\sdk\\ndk\\29.0.13846066"
ndkPath "D:\\Android\\sdk\\ndk\\27.0.12077973"
compileSdkVersion 35
buildToolsVersion '34.0.0'

BIN
unityLibrary/src/main/jniLibs/arm64-v8a/libil2cpp.so
View File

Binary file not shown.

BIN
unityLibrary/src/main/jniLibs/armeabi-v7a/libil2cpp.so
View File

Binary file not shown.

BIN
unityLibrary/symbols/arm64-v8a/libil2cpp.so
View File

Binary file not shown.

BIN
unityLibrary/symbols/armeabi-v7a/libil2cpp.so
View File

Binary file not shown.