Spaces:
Running
Running
| <html lang="en"> | |
| <head> | |
| <meta charset="UTF-8"> | |
| <meta name="viewport" content="width=device-width, initial-scale=1.0"> | |
| <title>Indic Audio Tools</title> | |
| <script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/vue.min.js"></script> | |
| <script src="https://cdn.jsdelivr.net/npm/axios/dist/axios.min.js"></script> | |
| <script src="https://cdn.jsdelivr.net/npm/@sanskrit-coders/[email protected]/sanscript.min.js"></script> | |
| <link href="https://unpkg.com/tailwindcss@^2/dist/tailwind.min.css" rel="stylesheet"> | |
| <link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.3/css/all.min.css" rel="stylesheet"> | |
| <link rel="stylesheet" href="https://cdn.plyr.io/3.6.8/plyr.css" /> | |
| <script src="https://cdn.plyr.io/3.6.8/plyr.polyfilled.js"></script> | |
| </head> | |
| <body class="bg-gray-100"> | |
| <div id="app" class="container mx-auto px-4 py-8"> | |
| <h1 class="text-3xl font-bold text-center mb-8 text-indigo-600">Indic Audio Tools</h1> | |
| <!-- Tabs --> | |
| <div class="flex justify-center mb-6"> | |
| <div class="inline-flex rounded-md shadow-sm" role="group"> | |
| <button @click="activeTab = 'transcription'" | |
| :class="{ 'bg-indigo-600 text-white': activeTab === 'transcription', 'bg-white text-indigo-600': activeTab !== 'transcription' }" | |
| class="px-4 py-2 text-sm font-medium border border-indigo-600 rounded-l-lg focus:z-10 focus:ring-2 focus:ring-indigo-500 focus:text-indigo-100 transition-colors duration-200"> | |
| Audio Transcription | |
| </button> | |
| <button @click="activeTab = 'tts'" | |
| :class="{ 'bg-indigo-600 text-white': activeTab === 'tts', 'bg-white text-indigo-600': activeTab !== 'tts' }" | |
| class="px-4 py-2 text-sm font-medium border border-indigo-600 rounded-r-lg focus:z-10 focus:ring-2 focus:ring-indigo-500 focus:text-indigo-100 transition-colors duration-200"> | |
| Text-to-Speech | |
| </button> | |
| </div> | |
| </div> | |
| <!-- Transcription Tab Content --> | |
| <div v-if="activeTab === 'transcription'" class="max-w-2xl mx-auto bg-white rounded-lg shadow-md p-6"> | |
| <!-- Transcription content here --> | |
| <div class="mb-4"> | |
| <label for="audioFile" class="block text-sm font-medium text-gray-700 mb-2">Select Audio File</label> | |
| <input type="file" id="audioFile" @change="handleFileChange" accept="audio/*" class="w-full px-3 py-2 border border-gray-300 rounded-md shadow-sm focus:outline-none focus:ring-indigo-500 focus:border-indigo-500"> | |
| </div> | |
| <div class="mb-4 flex items-center"> | |
| <button @click="toggleRecording" class="mr-4 px-4 py-2 bg-red-500 text-white rounded-md hover:bg-green-600 focus:outline-none focus:ring-2 focus:ring-offset-2 focus:ring-red-500" :class="{ 'bg-gray-600 hover:bg-gray-700': isRecording }"> | |
| <i class="fas" :class="isRecording ? 'fa-stop' : 'fa-microphone'"></i> | |
| {{ isRecording ? 'Stop Recording' : 'Record Audio' }} | |
| </button> | |
| <span v-if="isRecording" class="text-red-500">Recording... {{ recordingTime }}s</span> | |
| <span v-else-if="audioBlob" class="text-green-600">Recording complete ({{ recordingTime }}s)</span> | |
| </div> | |
| <div class="flex justify-center" v-if="audioFile"> | |
| <button @click="transcribeAudio" class="px-4 py-2 bg-indigo-600 text-white rounded-md hover:bg-indigo-700 focus:outline-none focus:ring-2 focus:ring-offset-2 focus:ring-indigo-500" :disabled="isLoading"> | |
| <span v-if="!isLoading">Transcribe Uploaded Audio</span> | |
| <span v-else class="flex items-center"> | |
| <svg class="animate-spin -ml-1 mr-3 h-5 w-5 text-white" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24"> | |
| <circle class="opacity-25" cx="12" cy="12" r="10" stroke="currentColor" stroke-width="4"></circle> | |
| <path class="opacity-75" fill="currentColor" d="M4 12a8 8 0 018-8V0C5.373 0 0 5.373 0 12h4zm2 5.291A7.962 7.962 0 014 12H0c0 3.042 1.135 5.824 3 7.938l3-2.647z"></path> | |
| </svg> | |
| Transcribing... | |
| </span> | |
| </button> | |
| </div> | |
| <div v-if="transcription" class="mt-6 bg-gray-100 p-4 rounded-lg"> | |
| <h3 class="text-lg font-semibold text-gray-700 mb-2">Transcription Result</h3> | |
| <p class="text-gray-600">{{ transcription }}</p> | |
| </div> | |
| </div> | |
| <!-- Text-to-Speech Tab Content --> | |
| <div v-if="activeTab === 'tts'" class="max-w-2xl mx-auto bg-white rounded-lg shadow-md p-6"> | |
| <!-- TTS content here --> | |
| <div class="mb-4"> | |
| <label for="language" class="block text-sm font-medium text-gray-700 mb-2">Select Language</label> | |
| <select id="language" v-model="selectedLanguage" @change="updateExample" class="w-full px-3 py-2 border border-gray-300 rounded-md shadow-sm focus:outline-none focus:ring-indigo-500 focus:border-indigo-500"> | |
| <option v-for="lang in languages" :value="lang.code">{{ lang.name }}</option> | |
| <option value="san">Sanskrit (Devanagari)</option> | |
| </select> | |
| </div> | |
| <div class="mb-4"> | |
| <label for="inputText" class="block text-sm font-medium text-gray-700 mb-2">Enter Text</label> | |
| <textarea id="inputText" v-model="inputText" rows="4" class="w-full px-3 py-2 border border-gray-300 rounded-md shadow-sm focus:outline-none focus:ring-indigo-500 focus:border-indigo-500" :placeholder="currentExample"></textarea> | |
| </div> | |
| <div class="flex justify-center"> | |
| <button @click="generateSpeech" class="px-4 py-2 bg-indigo-600 text-white rounded-md hover:bg-indigo-700 focus:outline-none focus:ring-2 focus:ring-offset-2 focus:ring-indigo-500" :disabled="isLoading"> | |
| <span v-if="!isLoading">Generate Speech</span> | |
| <span v-else class="flex items-center"> | |
| <svg class="animate-spin -ml-1 mr-3 h-5 w-5 text-white" xmlns="http://www.w3.org/2000/svg" fill="none" viewBox="0 0 24 24"> | |
| <circle class="opacity-25" cx="12" cy="12" r="10" stroke="currentColor" stroke-width="4"></circle> | |
| <path class="opacity-75" fill="currentColor" d="M4 12a8 8 0 018-8V0C5.373 0 0 5.373 0 12h4zm2 5.291A7.962 7.962 0 014 12H0c0 3.042 1.135 5.824 3 7.938l3-2.647z"></path> | |
| </svg> | |
| Generating... | |
| </span> | |
| </button> | |
| </div> | |
| <div v-if="audioUrl" class="mt-6 bg-gray-100 p-4 rounded-lg"> | |
| <div class="flex items-center justify-between mb-2"> | |
| <h3 class="text-lg font-semibold text-gray-700">Generated Audio</h3> | |
| <button @click="saveAudio" class="px-3 py-1 bg-green-600 text-white text-sm rounded-md hover:bg-green-700 focus:outline-none focus:ring-2 focus:ring-offset-2 focus:ring-green-500"> | |
| Save Audio | |
| </button> | |
| </div> | |
| <audio ref="audioPlayer" :src="audioUrl"></audio> | |
| </div> | |
| </div> | |
| </div> | |
| <script> | |
| new Vue({ | |
| el: '#app', | |
| data: { | |
| activeTab: 'transcription', | |
| // Transcription data | |
| audioFile: null, | |
| audioBlob: null, | |
| transcription: '', | |
| isLoading: false, | |
| isRecording: false, | |
| mediaRecorder: null, | |
| audioChunks: [], | |
| recordingTime: 0, | |
| recordingInterval: null, | |
| // TTS data | |
| languages: [], | |
| selectedLanguage: '', | |
| inputText: '', | |
| audioUrl: null, | |
| player: null, | |
| examples: { | |
| hin: "भारत विविधता में एकता का देश है। यहाँ अनेक भाषाएँ, धर्म और संस्कृतियाँ एक साथ फलती-फूलती हैं।", | |
| ben: "বাংলা সাহিত্যের ঐতিহ্য বিশ্বব্যাপী স্বীকৃত। রবীন্দ্রনাথ ঠাকুর এর একজন মহান প্রতিনিধি।", | |
| mar: "महाराष्ट्र हे भारतातील एक प्रगत राज्य आहे. मुंबई ही या राज्याची राजधानी आणि आर्थिक केंद्र आहे।", | |
| tel: "తెలుగు భాష మధురమైనది మరియు సంగీతాత్మకమైనది. ఆంధ్ర ప్రదేశ్ మరియు తెలంగాణ రాష్ట్రాల అధికార భాష ఇదే.", | |
| tam: "தமிழ் மொழி மிகவும் பழமையான மொழிகளில் ஒன்றாகும். இது தென்னிந்தியாவில் பரவலாகப் பேசப்படுகிறது.", | |
| guj: "ગુજરાત એ ભારતનું એક સમૃદ્ધ રાજ્ય છે. અહીં આધુનિકતા અને પરંપરા સાથે ચાલે છે.", | |
| "urd-script_arabic": "اردو ایک خوبصورت زبان ہے جو پاکستان اور بھارت میں بولی جاتی ہے۔ اس کی شاعری دنیا بھر میں مشہور ہے۔", | |
| kan: "ಕರ್ನಾಟಕ ರಾಜ್ಯವು ಅದ್ಭುತ ಸಂಸ್ಕೃತಿ ಮತ್ತು ಪ್ರಕೃತಿ ಸೌಂದರ್ಯಕ್ಕೆ ಹೆಸರುವಾಸಿಯಾಗಿದೆ. ಇದರ ರಾಜಧಾನಿ ಬೆಂಗಳೂರು ಭಾರತದ ತಂತ್ರಜ್ಞಾನ ಕೇಂದ್ರವಾಗಿದೆ.", | |
| mal: "കേരളം ഇന്ത്യയുടെ തെക്കുപടിഞ്ഞാറൻ സംസ്ഥാനമാണ്. ഇവിടെ പ്രകൃതിയുടെ സൗന്ദര്യവും സമൃദ്ധമായ സംസ്കാരവും കാണാം.", | |
| pan: "ਪੰਜਾਬ ਦੀ ਧਰਤੀ ਬਹੁਤ ਉਪਜਾਊ ਹੈ। ਇੱਥੇ ਦੀ ਖੇਤੀਬਾੜੀ ਅਤੇ ਸੱਭਿਆਚਾਰ ਪੂਰੇ ਭਾਰਤ ਵਿੱਚ ਪ੍ਰਸਿੱਧ ਹਨ।", | |
| san: "संस्कृतं भारतस्य प्राचीनतमा भाषा अस्ति। इयं देवभाषा इति अपि कथ्यते। अस्याः साहित्यं विपुलं वैविध्यपूर्णं च विद्यते।" | |
| }, | |
| currentExample: '' | |
| }, | |
| mounted() { | |
| this.fetchLanguages(); | |
| }, | |
| updated() { | |
| this.$nextTick(() => { | |
| if (this.audioUrl && this.$refs.audioPlayer) { | |
| if (this.player) { | |
| this.player.destroy(); | |
| } | |
| this.player = new Plyr(this.$refs.audioPlayer, { | |
| controls: ['play', 'progress', 'current-time', 'mute', 'volume'] | |
| }); | |
| } | |
| }); | |
| }, | |
| methods: { | |
| // Transcription methods | |
| handleFileChange(event) { | |
| this.audioFile = event.target.files[0]; | |
| this.audioBlob = null; | |
| this.transcription = ''; | |
| }, | |
| async toggleRecording() { | |
| if (this.isRecording) { | |
| this.stopRecording(); | |
| } else { | |
| await this.startRecording(); | |
| } | |
| }, | |
| async startRecording() { | |
| try { | |
| const stream = await navigator.mediaDevices.getUserMedia({ audio: true }); | |
| this.mediaRecorder = new MediaRecorder(stream); | |
| this.audioChunks = []; | |
| this.mediaRecorder.ondataavailable = (event) => { | |
| this.audioChunks.push(event.data); | |
| }; | |
| this.mediaRecorder.onstop = () => { | |
| this.audioBlob = new Blob(this.audioChunks, { type: 'audio/wav' }); | |
| this.audioFile = null; | |
| this.transcribeAudio(); // Automatically transcribe after recording stops | |
| }; | |
| this.mediaRecorder.start(); | |
| this.isRecording = true; | |
| this.recordingTime = 0; | |
| this.recordingInterval = setInterval(() => { | |
| this.recordingTime++; | |
| }, 1000); | |
| } catch (error) { | |
| console.error('Error accessing microphone:', error); | |
| alert('Unable to access the microphone. Please make sure it\'s connected and you\'ve granted permission.'); | |
| } | |
| }, | |
| stopRecording() { | |
| if (this.mediaRecorder) { | |
| this.mediaRecorder.stop(); | |
| this.isRecording = false; | |
| clearInterval(this.recordingInterval); | |
| } | |
| }, | |
| transcribeAudio() { | |
| if (!this.audioFile && !this.audioBlob) { | |
| alert('Please select an audio file or record audio'); | |
| return; | |
| } | |
| this.isLoading = true; | |
| const formData = new FormData(); | |
| if (this.audioFile) { | |
| formData.append('file', this.audioFile); | |
| } else { | |
| formData.append('file', this.audioBlob, 'recorded_audio.wav'); | |
| } | |
| axios.post('https://pvanand-audio-chat-indic.hf.space/transcribe/', formData, { | |
| headers: { | |
| 'Content-Type': 'multipart/form-data' | |
| } | |
| }) | |
| .then(response => { | |
| this.transcription = response.data.transcript; | |
| this.isLoading = false; | |
| }) | |
| .catch(error => { | |
| console.error('Error transcribing audio:', error); | |
| alert('Error transcribing audio. Please try again.'); | |
| this.isLoading = false; | |
| }); | |
| }, | |
| // TTS methods | |
| fetchLanguages() { | |
| axios.get('https://pvanand-audio-chat-indic.hf.space/languages') | |
| .then(response => { | |
| this.languages = response.data; | |
| if (this.languages.length > 0) { | |
| this.selectedLanguage = this.languages[0].code; | |
| this.updateExample(); | |
| } | |
| }) | |
| .catch(error => console.error('Error fetching languages:', error)); | |
| }, | |
| updateExample() { | |
| this.currentExample = this.examples[this.selectedLanguage] || ''; | |
| this.inputText = ''; | |
| this.audioUrl = null; | |
| }, | |
| generateSpeech() { | |
| this.isLoading = true; | |
| let text = this.inputText.trim() || this.currentExample; | |
| let lang = this.selectedLanguage; | |
| if (this.selectedLanguage === 'san') { | |
| if (typeof Sanscript !== 'undefined') { | |
| text = Sanscript.t(text, 'devanagari', 'kannada'); | |
| lang = 'kan'; // Use Kannada endpoint for Sanskrit | |
| } else { | |
| console.error('Sanscript library not available'); | |
| alert('Sanskrit conversion is not available at the moment.'); | |
| this.isLoading = false; | |
| return; | |
| } | |
| } | |
| const url = `https://pvanand-audio-chat-indic.hf.space/tts?text=${encodeURIComponent(text)}&lang=${lang}`; | |
| axios.get(url, { responseType: 'blob' }) | |
| .then(response => { | |
| const blob = new Blob([response.data], { type: 'audio/mp3' }); | |
| this.audioUrl = URL.createObjectURL(blob); | |
| this.isLoading = false; | |
| }) | |
| .catch(error => { | |
| console.error('Error generating speech:', error); | |
| this.isLoading = false; | |
| }); | |
| }, | |
| saveAudio() { | |
| const link = document.createElement('a'); | |
| link.href = this.audioUrl; | |
| link.download = 'generated_speech.mp3'; | |
| document.body.appendChild(link); | |
| link.click(); | |
| document.body.removeChild(link); | |
| } | |
| } | |
| }); | |
| </script> | |
| </body> | |
| </html> |