init

services/a2f_api/README.md

@@ -0,0 +1,64 @@
+# Text to BlendShapes API
+
+将文字转换为52个形态键的API服务
+
+## 功能
+
+1. 文字 → 音频 (使用 gTTS)
+2. 音频 → CSV (使用 Audio2Face)
+3. CSV → 52个形态键数据
+
+## 安装
+
+```bash
+cd a2f_api
+pip install -r requirements.txt
+```
+
+## 使用
+
+```bash
+python api.py
+```
+
+服务将在 `http://localhost:5001` 启动
+
+## API 接口
+
+### POST /text-to-blendshapes
+
+**请求:**
+```json
+{
+  "text": "你好世界",
+  "language": "zh-CN"
+}
+```
+
+**响应:**
+```json
+{
+  "success": true,
+  "frames": [
+    {
+      "timeCode": 0.0,
+      "blendShapes": {
+        "EyeBlinkLeft": 0.0,
+        "EyeLookDownLeft": 0.0,
+        ...
+        "TongueOut": 0.0
+      }
+    }
+  ],
+  "audio_path": "...",
+  "csv_path": "..."
+}
+```
+
+## 文件说明
+
+- `tts_service.py` - 文字转音频服务
+- `a2f_service.py` - Audio2Face包装器
+- `blend_shape_parser.py` - CSV解析器
+- `text_to_blendshapes_service.py` - 主服务
+- `api.py` - Flask API

services/a2f_api/a2f_service.py

@@ -0,0 +1,40 @@
+import subprocess
+import sys
+import os
+from pathlib import Path
+import glob
+
+class A2FService:
+    def __init__(self, a2f_url="192.168.1.39:52000"):
+        self.base_dir = Path(__file__).parent.parent.parent
+        self.output_dir = self.base_dir / "data" / "output"
+        self.a2f_script = self.base_dir / "external" / "Audio2Face-3D-Samples" / "scripts" / "audio2face_3d_microservices_interaction_app" / "a2f_3d.py"
+        self.config_file = self.base_dir / "external" / "Audio2Face-3D-Samples" / "scripts" / "audio2face_3d_microservices_interaction_app" / "config" / "config_james.yml"
+        self.a2f_url = a2f_url
+        os.makedirs(self.output_dir, exist_ok=True)
+
+    def audio_to_csv(self, audio_path: str) -> str:
+        cmd = [
+            sys.executable,
+            str(self.a2f_script),
+            "run_inference",
+            audio_path,
+            str(self.config_file),
+            "--url",
+            self.a2f_url
+        ]
+
+        result = subprocess.run(cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, text=True, cwd=str(self.output_dir))
+
+        if result.returncode != 0:
+            raise RuntimeError(f"A2F inference failed: {result.stdout}")
+
+        output_dirs = sorted(glob.glob(str(self.output_dir / "output_*")))
+        if not output_dirs:
+            raise RuntimeError("No output directory found")
+
+        csv_path = os.path.join(output_dirs[-1], "animation_frames.csv")
+        if not os.path.exists(csv_path):
+            raise RuntimeError(f"CSV file not found: {csv_path}")
+
+        return csv_path

services/a2f_api/api.py

@@ -0,0 +1,34 @@
+from flask import Flask, request, jsonify
+from flask_cors import CORS
+from text_to_blendshapes_service import TextToBlendShapesService
+
+app = Flask(__name__)
+CORS(app)
+
+@app.route('/health', methods=['GET'])
+def health():
+    return jsonify({'status': 'ok'})
+
+@app.route('/text-to-blendshapes', methods=['POST'])
+def text_to_blendshapes():
+    try:
+        data = request.get_json()
+        if not data or 'text' not in data:
+            return jsonify({'success': False, 'error': 'Missing text'}), 400
+
+        text = data['text']
+        language = data.get('language', 'zh-CN')
+
+        service = TextToBlendShapesService(lang=language)
+        result = service.text_to_blend_shapes(text)
+
+        return jsonify(result)
+
+    except Exception as e:
+        import traceback
+        traceback.print_exc()
+        return jsonify({'success': False, 'error': str(e)}), 500
+
+if __name__ == '__main__':
+    print("Text to BlendShapes API: http://localhost:5001")
+    app.run(host='0.0.0.0', port=5001, debug=True)

services/a2f_api/blend_shape_parser.py

@@ -0,0 +1,31 @@
+import csv
+
+class BlendShapeParser:
+    BLEND_SHAPE_KEYS = [
+        'EyeBlinkLeft', 'EyeLookDownLeft', 'EyeLookInLeft', 'EyeLookOutLeft', 'EyeLookUpLeft',
+        'EyeSquintLeft', 'EyeWideLeft', 'EyeBlinkRight', 'EyeLookDownRight', 'EyeLookInRight',
+        'EyeLookOutRight', 'EyeLookUpRight', 'EyeSquintRight', 'EyeWideRight', 'JawForward',
+        'JawLeft', 'JawRight', 'JawOpen', 'MouthClose', 'MouthFunnel', 'MouthPucker',
+        'MouthLeft', 'MouthRight', 'MouthSmileLeft', 'MouthSmileRight', 'MouthFrownLeft',
+        'MouthFrownRight', 'MouthDimpleLeft', 'MouthDimpleRight', 'MouthStretchLeft',
+        'MouthStretchRight', 'MouthRollLower', 'MouthRollUpper', 'MouthShrugLower',
+        'MouthShrugUpper', 'MouthPressLeft', 'MouthPressRight', 'MouthLowerDownLeft',
+        'MouthLowerDownRight', 'MouthUpperUpLeft', 'MouthUpperUpRight', 'BrowDownLeft',
+        'BrowDownRight', 'BrowInnerUp', 'BrowOuterUpLeft', 'BrowOuterUpRight', 'CheekPuff',
+        'CheekSquintLeft', 'CheekSquintRight', 'NoseSneerLeft', 'NoseSneerRight', 'TongueOut'
+    ]
+
+    @staticmethod
+    def csv_to_blend_shapes(csv_path: str):
+        frames = []
+        with open(csv_path, 'r') as f:
+            
+            reader = csv.DictReader(f)
+            for row in reader:
+                frame = {'timeCode': float(row['timeCode']), 'blendShapes': {}}
+                for key in BlendShapeParser.BLEND_SHAPE_KEYS:
+                    col_name = f'blendShapes.{key}'
+                    if col_name in row:
+                        frame['blendShapes'][key] = float(row[col_name])
+                frames.append(frame)
+        return frames

services/a2f_api/requirements.txt

@@ -0,0 +1,3 @@
+flask>=3.0.0
+flask-cors>=4.0.0
+pyttsx3>=2.90

services/a2f_api/text_to_blendshapes_service.py

@@ -0,0 +1,31 @@
+import os
+import tempfile
+from datetime import datetime
+from tts_service import TTSService
+from a2f_service import A2FService
+from blend_shape_parser import BlendShapeParser
+
+class TextToBlendShapesService:
+    def __init__(self, lang='zh-CN', a2f_url="192.168.1.39:52000"):
+        self.tts = TTSService(lang=lang)
+        self.a2f = A2FService(a2f_url=a2f_url)
+        self.parser = BlendShapeParser()
+
+    def text_to_blend_shapes(self, text: str, output_dir: str = None):
+        if output_dir is None:
+            output_dir = tempfile.gettempdir()
+
+        os.makedirs(output_dir, exist_ok=True)
+        timestamp = datetime.now().strftime('%Y%m%d%H%M%S')
+        audio_path = os.path.join(output_dir, f'tts_{timestamp}.wav')
+
+        self.tts.text_to_audio(text, audio_path)
+        csv_path = self.a2f.audio_to_csv(audio_path)
+        frames = self.parser.csv_to_blend_shapes(csv_path)
+
+        return {
+            'success': True,
+            'frames': frames,
+            'audio_path': audio_path,
+            'csv_path': csv_path
+        }

services/a2f_api/tts_service.py

@@ -0,0 +1,20 @@
+import os
+import pyttsx3
+
+class TTSService:
+    def __init__(self, lang='zh-CN'):
+        self.lang = lang
+        self.engine = pyttsx3.init()
+
+        if lang == 'zh-CN':
+            voices = self.engine.getProperty('voices')
+            for voice in voices:
+                if 'chinese' in voice.name.lower() or 'zh' in voice.id.lower():
+                    self.engine.setProperty('voice', voice.id)
+                    break
+
+    def text_to_audio(self, text: str, output_path: str) -> str:
+        os.makedirs(os.path.dirname(output_path), exist_ok=True)
+        self.engine.save_to_file(text, output_path)
+        self.engine.runAndWait()
+        return output_path