integrated simli

examples/foundational/25-simli-layer.py

@@ -0,0 +1,141 @@
+import asyncio
+import aiohttp
+import os
+import sys
+
+from pipecat.audio.vad.silero import SileroVADAnalyzer
+from pipecat.pipeline.pipeline import Pipeline
+from pipecat.pipeline.runner import PipelineRunner
+from pipecat.pipeline.task import PipelineParams, PipelineTask
+from pipecat.frames.frames import LLMMessagesFrame
+
+from pipecat.processors.aggregators.openai_llm_context import OpenAILLMContext
+from pipecat.services.ai_services import AIService
+from pipecat.services.cartesia import CartesiaTTSService
+from pipecat.services.elevenlabs import ElevenLabsTTSService
+from pipecat.services.openai import OpenAILLMService
+from pipecat.transports.services.daily import DailyParams, DailyTransport
+
+from pipecat.transports.services.helpers.daily_rest import DailyRESTHelper, DailyRoomParams
+from runner import configure
+from loguru import logger
+from dotenv import load_dotenv
+
+from simli import SimliConfig
+from pipecat.services.simli import SimliVideoService
+
+load_dotenv(override=True)
+
+logger.remove(0)
+logger.add(sys.stderr, level="DEBUG")
+
+
+async def main():
+    async with aiohttp.ClientSession() as session:
+        (room_url, token) = await configure(session)
+        
+        print("Creating room")
+        aiohttp_session = aiohttp.ClientSession()
+        daily_helper = DailyRESTHelper(
+            daily_api_key=os.getenv("DAILY_API_KEY", ""),
+            daily_api_url=os.getenv("DAILY_API_URL", "https://api.daily.co/v1"),
+            aiohttp_session=aiohttp_session,
+        )
+        
+        room = await daily_helper.create_room(DailyRoomParams())
+        expiry_time: float = 60 * 60
+
+        token = await daily_helper.get_token(room.url, expiry_time)
+
+        
+        print("Room created ", room.url)
+        
+        transport = DailyTransport(
+            room.url,
+            token,
+            "Chatbot",
+            DailyParams(
+                audio_out_enabled=True,
+                camera_out_enabled=True,
+                camera_out_width=512,
+                camera_out_height=512,
+                vad_enabled=True,
+                vad_analyzer=SileroVADAnalyzer(),
+                transcription_enabled=True,
+                #
+                # Spanish
+                #
+                # transcription_settings=DailyTranscriptionSettings(
+                #     language="es",
+                #     tier="nova",
+                #     model="2-general"
+                # )
+            ),
+        )
+
+        # tts = ElevenLabsTTSService(
+        #     api_key=os.getenv("ELEVENLABS_API_KEY"),
+        #     voice_id="pNInz6obpgDQGcFmaJgB",
+        # )
+
+        tts = CartesiaTTSService(
+            api_key=os.getenv("CARTESIA"),
+            voice_id="a167e0f3-df7e-4d52-a9c3-f949145efdab",
+        )
+
+        llm = OpenAILLMService(api_key=os.getenv("OPENAI_API_KEY"), model="gpt-4o-mini")
+        
+        messages = [
+            {
+                "role": "system",
+                #
+                # English
+                #
+                "content": "You are a helpful LLM in a WebRTC call. Your goal is to demonstrate your capabilities in a succinct way. Your output will be converted to audio so don't include special characters in your answers. Respond to what the user said in a creative and helpful way.",
+                #
+                # Spanish
+                #
+                # "content": "Eres Chatbot, un amigable y útil robot. Tu objetivo es demostrar tus capacidades de una manera breve. Tus respuestas se convertiran a audio así que nunca no debes incluir caracteres especiales. Contesta a lo que el usuario pregunte de una manera creativa, útil y breve. Empieza por presentarte a ti mismo.",
+            },
+        ]
+        simliAi = SimliVideoService(
+            SimliConfig(os.getenv("SIMLI_API_KEY"), os.getenv("SIMLI_FACE_ID"))
+        )
+        print("starting connection to simi")
+        await simliAi.startConnection()
+        print("connection started")
+
+        context = OpenAILLMContext(messages)
+        context_aggregator = llm.create_context_aggregator(context)
+
+        pipeline = Pipeline(
+            [
+                transport.input(),
+                context_aggregator.user(),
+                llm,
+                tts,
+                simliAi,
+                transport.output(),
+                context_aggregator.assistant(),
+            ]
+        )
+
+        task = PipelineTask(
+            pipeline,
+            PipelineParams(
+                allow_interruptions=True,
+                enable_metrics=True,
+            ),
+        )
+
+        @transport.event_handler("on_first_participant_joined")
+        async def on_first_participant_joined(transport, participant):
+            await transport.capture_participant_transcription(participant["id"])
+            await task.queue_frames([LLMMessagesFrame(messages)])
+
+        runner = PipelineRunner()
+        await runner.run(task)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

pyproject.toml

@@ -66,7 +66,7 @@ soundfile = [ "soundfile~=0.12.1" ]
 together = [ "openai~=1.50.2" ]
 websocket = [ "websockets~=13.1", "fastapi~=0.115.0" ]
 whisper = [ "faster-whisper~=1.0.3" ]
-
+simli = [ "simli-ai>=0.1.7"]
 [tool.setuptools.packages.find]
 # All the following settings are optional:
 where = ["src"]

src/pipecat/services/simli.py

@@ -0,0 +1,116 @@
+import asyncio
+
+from pipecat.frames.frames import (
+    Frame,
+    OutputImageRawFrame,
+    TTSAudioRawFrame,
+    StartInterruptionFrame,
+    EndFrame,
+    CancelFrame,
+)
+from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
+
+import numpy as np
+from av import AudioFrame
+from av.audio.resampler import AudioResampler
+from simli import SimliClient, SimliConfig
+
+
+class SimliVideoService(FrameProcessor):
+    def __init__(
+        self, simliConfig: SimliConfig, useTurnServer=False, latencyInterval=60
+    ):
+        super().__init__()
+        self.simliClient = SimliClient(simliConfig, useTurnServer, latencyInterval)
+
+        self.pipecatResampler: AudioResampler = None
+        self.name = "SimliAi"
+        self.ready = False
+        self.simliResampler = AudioResampler("s16", 1, 16000)
+        self.AudioTask: asyncio.Task = None
+        self.VideoTask: asyncio.Task = None
+
+    async def startConnection(self):
+        await self.simliClient.Initialize()
+        self.ready = True
+        # Create task to consume and process audio and video
+        self.AudioTask = asyncio.create_task(self.consume_and_process_audio())
+        self.VideoTask = asyncio.create_task(self.consume_and_process_video())
+
+    async def consume_and_process_audio(self):
+        async for audio_frame in self.simliClient.getAudioStreamIterator():
+            # Process the audio frame
+            try:
+                resampledFrames = self.pipecatResampler.resample(audio_frame)
+                for resampled_frame in resampledFrames:
+                    await self.push_frame(
+                        TTSAudioRawFrame(
+                            audio=resampled_frame.to_ndarray().tobytes(),
+                            sample_rate=self.pipecatResampler.rate,
+                            num_channels=1,
+                        ),
+                    )
+            except Exception as e:
+                print(e)
+                import traceback
+
+                traceback.print_exc()
+
+    async def consume_and_process_video(self):
+        async for video_frame in self.simliClient.getVideoStreamIterator(
+            targetFormat="rgb24"
+        ):
+            # Process the video frame
+            convertedFrame: OutputImageRawFrame = OutputImageRawFrame(
+                image=video_frame.to_rgb().to_image().tobytes(),
+                size=(video_frame.width, video_frame.height),
+                format="RGB",
+            )
+            convertedFrame.pts = video_frame.pts
+            await self.push_frame(
+                convertedFrame,
+            )
+
+    async def process_frame(self, frame: Frame, direction: FrameDirection):
+        await super().process_frame(frame, direction)
+
+        if isinstance(frame, TTSAudioRawFrame):
+            # Send audio frame to Simli
+            try:
+                if self.ready:
+                    AudioFrame
+                    oldFrame = AudioFrame.from_ndarray(
+                        np.frombuffer(frame.audio, dtype=np.int16)[None, :],
+                        layout=frame.num_channels,
+                    )
+                    oldFrame.sample_rate = frame.sample_rate
+                    if self.pipecatResampler is None:
+                        self.pipecatResampler = AudioResampler(
+                            "s16", oldFrame.layout, oldFrame.sample_rate
+                        )
+
+                    resampledFrame = self.simliResampler.resample(oldFrame)
+                    for frame in resampledFrame:
+                        await self.simliClient.send(
+                            frame.to_ndarray().astype(np.int16).tobytes()
+                        )
+                    return
+                else:
+                    print(
+                        "Simli Connection is not Initialized properly, passing audio to next processor"
+                    )
+                    await self.push_frame(frame, direction)
+            except Exception as e:
+                print(e)
+                import traceback
+
+                traceback.print_exc()
+        elif isinstance(frame, (EndFrame, CancelFrame)):
+            await self.simliClient.stop()
+            self.AudioTask.cancel()
+            self.VideoTask.cancel()
+
+        elif isinstance(frame, StartInterruptionFrame):
+            await self.simliClient.clearBuffer()
+
+        await self.push_frame(frame, direction)