Reorganize examples into topic-based subfolders
Move 304 examples from a flat numbered directory into 14 descriptive subfolders: getting-started, services (speech + function-calling), transcription, vision, realtime, persistent-context, context-summarization, update-settings (stt/tts/llm), turn-management, thinking-and-mcp, transports, video-avatar, video-processing, and features. Strip numbered prefixes from filenames (e.g. 07c-interruptible-deepgram.py becomes services/speech/deepgram.py) since the folder context makes them redundant. Keep numbered prefixes only in getting-started/ where ordering matters. Update eval script paths and README to match the new structure.
This commit is contained in:
Mark Backman
208
examples/video-processing/custom-video-track.py
Normal file
208
examples/video-processing/custom-video-track.py
Normal file
@@ -0,0 +1,208 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
"""Example demonstrating custom video tracks output with Daily transport.
|
||||
|
||||
This example outputs two video track simultaneously:
|
||||
- The default camera track with an animated color gradient pattern.
|
||||
- A custom "blue" track with the same pattern but with a blue tint applied.
|
||||
|
||||
The pattern generator pushes frames to the default camera. A second processor
|
||||
(BlueTintProcessor) duplicates each frame, applies a blue tint, and pushes it
|
||||
to the "blue" custom video destination.
|
||||
"""
|
||||
|
||||
import asyncio
|
||||
import math
|
||||
import time
|
||||
|
||||
import numpy as np
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.frames.frames import (
|
||||
CancelFrame,
|
||||
EndFrame,
|
||||
Frame,
|
||||
OutputImageRawFrame,
|
||||
StartFrame,
|
||||
SystemFrame,
|
||||
)
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineTask
|
||||
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.transports.base_transport import BaseTransport
|
||||
from pipecat.transports.daily.transport import DailyCustomVideoTrackParams, DailyParams
|
||||
|
||||
WIDTH = 320
|
||||
HEIGHT = 240
|
||||
FPS = 30
|
||||
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
video_out_enabled=True,
|
||||
video_out_width=WIDTH,
|
||||
video_out_height=HEIGHT,
|
||||
video_out_framerate=FPS,
|
||||
video_out_destinations=["blue"],
|
||||
custom_video_track_params={
|
||||
"blue": DailyCustomVideoTrackParams(
|
||||
width=WIDTH,
|
||||
height=HEIGHT,
|
||||
send_settings={
|
||||
"maxQuality": "low",
|
||||
"encodings": {
|
||||
"low": {
|
||||
"maxBitrate": 500_000,
|
||||
"maxFramerate": FPS,
|
||||
}
|
||||
},
|
||||
},
|
||||
),
|
||||
},
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
def generate_gradient_frame(width: int, height: int, t: float) -> np.ndarray:
|
||||
"""Generate an animated gradient pattern.
|
||||
|
||||
Creates a smooth color gradient that shifts over time using sine waves
|
||||
for each RGB channel at different frequencies.
|
||||
"""
|
||||
x = np.linspace(0, 1, width)
|
||||
y = np.linspace(0, 1, height)
|
||||
xv, yv = np.meshgrid(x, y)
|
||||
|
||||
r = ((np.sin(2 * math.pi * (xv + t * 0.3)) + 1) / 2 * 255).astype(np.uint8)
|
||||
g = ((np.sin(2 * math.pi * (yv + t * 0.5)) + 1) / 2 * 255).astype(np.uint8)
|
||||
b = ((np.sin(2 * math.pi * (xv + yv + t * 0.7)) + 1) / 2 * 255).astype(np.uint8)
|
||||
|
||||
return np.stack([r, g, b], axis=-1)
|
||||
|
||||
|
||||
class VideoPatternGenerator(FrameProcessor):
|
||||
"""Generates an animated gradient pattern and pushes it as video frames."""
|
||||
|
||||
def __init__(self, width: int, height: int, fps: int):
|
||||
super().__init__()
|
||||
self._width = width
|
||||
self._height = height
|
||||
self._fps = fps
|
||||
self._generate_task = None
|
||||
|
||||
async def process_frame(self, frame: Frame, direction: FrameDirection):
|
||||
await super().process_frame(frame, direction)
|
||||
|
||||
if isinstance(frame, StartFrame):
|
||||
await self.push_frame(frame, direction)
|
||||
await self._start()
|
||||
elif isinstance(frame, (EndFrame, CancelFrame)):
|
||||
await self._stop()
|
||||
await self.push_frame(frame, direction)
|
||||
else:
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
async def _start(self):
|
||||
self._generate_task = self.create_task(self._generate_loop(), "video_generate_loop")
|
||||
|
||||
async def _stop(self):
|
||||
if self._generate_task:
|
||||
await self.cancel_task(self._generate_task)
|
||||
self._generate_task = None
|
||||
|
||||
async def _generate_loop(self):
|
||||
interval = 1.0 / self._fps
|
||||
start = time.monotonic()
|
||||
|
||||
while True:
|
||||
t = time.monotonic() - start
|
||||
|
||||
pattern = generate_gradient_frame(self._width, self._height, t)
|
||||
|
||||
frame = OutputImageRawFrame(
|
||||
image=pattern.tobytes(),
|
||||
size=(self._width, self._height),
|
||||
format="RGB",
|
||||
)
|
||||
await self.push_frame(frame)
|
||||
|
||||
elapsed = time.monotonic() - start - t
|
||||
await asyncio.sleep(max(0, interval - elapsed))
|
||||
|
||||
|
||||
class BlueTintProcessor(FrameProcessor):
|
||||
"""Duplicates OutputImageRawFrames with a blue tint for a custom video destination."""
|
||||
|
||||
def __init__(self, destination: str):
|
||||
super().__init__()
|
||||
self._destination = destination
|
||||
|
||||
async def process_frame(self, frame: Frame, direction: FrameDirection):
|
||||
await super().process_frame(frame, direction)
|
||||
|
||||
if isinstance(frame, OutputImageRawFrame):
|
||||
# Pass through the original frame.
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
# Create a blue-tinted copy for the custom destination.
|
||||
img = np.frombuffer(frame.image, dtype=np.uint8).reshape(
|
||||
(frame.size[1], frame.size[0], 3)
|
||||
)
|
||||
tinted = img.copy()
|
||||
tinted[:, :, 0] = (tinted[:, :, 0] * 0.3).astype(np.uint8) # R
|
||||
tinted[:, :, 1] = (tinted[:, :, 1] * 0.3).astype(np.uint8) # G
|
||||
tinted[:, :, 2] = np.clip(tinted[:, :, 2].astype(np.uint16) + 80, 0, 255).astype(
|
||||
np.uint8
|
||||
) # B
|
||||
|
||||
blue_frame = OutputImageRawFrame(
|
||||
image=tinted.tobytes(),
|
||||
size=frame.size,
|
||||
format=frame.format,
|
||||
)
|
||||
blue_frame.transport_destination = self._destination
|
||||
await self.push_frame(blue_frame)
|
||||
else:
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
logger.info("Starting dual video track bot")
|
||||
|
||||
generator = VideoPatternGenerator(WIDTH, HEIGHT, FPS)
|
||||
blue_tint = BlueTintProcessor(destination="blue")
|
||||
|
||||
task = PipelineTask(
|
||||
Pipeline([generator, blue_tint, transport.output()]),
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
@transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, client):
|
||||
logger.info("Client connected")
|
||||
|
||||
@transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info("Client disconnected")
|
||||
await task.queue_frame(EndFrame())
|
||||
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
92
examples/video-processing/gstreamer-filesrc.py
Normal file
92
examples/video-processing/gstreamer-filesrc.py
Normal file
@@ -0,0 +1,92 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
import argparse
|
||||
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineTask
|
||||
from pipecat.processors.gstreamer.pipeline_source import GStreamerPipelineSource
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.transports.base_transport import BaseTransport, TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
# We use lambdas to defer transport parameter creation until the transport
|
||||
# type is selected at runtime.
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
audio_out_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
audio_out_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
logger.info(f"Starting bot with video input: {runner_args.cli_args.input}")
|
||||
|
||||
gst = GStreamerPipelineSource(
|
||||
pipeline=f"filesrc location={runner_args.cli_args.input}",
|
||||
out_params=GStreamerPipelineSource.OutputParams(
|
||||
video_width=1280,
|
||||
video_height=720,
|
||||
),
|
||||
)
|
||||
|
||||
pipeline = Pipeline(
|
||||
[
|
||||
gst, # GStreamer file source
|
||||
transport.output(), # Transport bot output
|
||||
]
|
||||
)
|
||||
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
@transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, client):
|
||||
logger.info(f"Client connected")
|
||||
|
||||
@transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info(f"Client disconnected")
|
||||
await task.cancel()
|
||||
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
parser = argparse.ArgumentParser(description="Pipecat Video Streaming Bot")
|
||||
parser.add_argument("-i", "--input", type=str, required=True, help="Input video file")
|
||||
|
||||
main(parser)
|
||||
76
examples/video-processing/gstreamer-videotestsrc.py
Normal file
76
examples/video-processing/gstreamer-videotestsrc.py
Normal file
@@ -0,0 +1,76 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineTask
|
||||
from pipecat.processors.gstreamer.pipeline_source import GStreamerPipelineSource
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.transports.base_transport import BaseTransport, TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
# We use lambdas to defer transport parameter creation until the transport
|
||||
# type is selected at runtime.
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
logger.info(f"Starting bot with video test source")
|
||||
|
||||
gst = GStreamerPipelineSource(
|
||||
pipeline='videotestsrc ! capsfilter caps="video/x-raw,width=1280,height=720,framerate=30/1"',
|
||||
out_params=GStreamerPipelineSource.OutputParams(
|
||||
video_width=1280, video_height=720, clock_sync=False
|
||||
),
|
||||
)
|
||||
|
||||
pipeline = Pipeline(
|
||||
[
|
||||
gst, # GStreamer test source
|
||||
transport.output(), # Transport bot output
|
||||
]
|
||||
)
|
||||
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
132
examples/video-processing/local-mirror.py
Normal file
132
examples/video-processing/local-mirror.py
Normal file
@@ -0,0 +1,132 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
import asyncio
|
||||
import tkinter as tk
|
||||
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.frames.frames import (
|
||||
Frame,
|
||||
InputAudioRawFrame,
|
||||
InputImageRawFrame,
|
||||
OutputAudioRawFrame,
|
||||
OutputImageRawFrame,
|
||||
)
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineParams, PipelineTask
|
||||
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport, maybe_capture_participant_camera
|
||||
from pipecat.transports.base_transport import BaseTransport, TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams
|
||||
from pipecat.transports.local.tk import TkLocalTransport, TkTransportParams
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
|
||||
class MirrorProcessor(FrameProcessor):
|
||||
async def process_frame(self, frame: Frame, direction: FrameDirection):
|
||||
await super().process_frame(frame, direction)
|
||||
|
||||
if isinstance(frame, InputAudioRawFrame):
|
||||
await self.push_frame(
|
||||
OutputAudioRawFrame(
|
||||
audio=frame.audio,
|
||||
sample_rate=frame.sample_rate,
|
||||
num_channels=frame.num_channels,
|
||||
)
|
||||
)
|
||||
elif isinstance(frame, InputImageRawFrame):
|
||||
await self.push_frame(
|
||||
OutputImageRawFrame(image=frame.image, size=frame.size, format=frame.format)
|
||||
)
|
||||
else:
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
|
||||
# We use lambdas to defer transport parameter creation until the transport
|
||||
# type is selected at runtime.
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
logger.info(f"Starting bot")
|
||||
|
||||
tk_root = tk.Tk()
|
||||
tk_root.title("Local Mirror")
|
||||
|
||||
tk_transport = TkLocalTransport(
|
||||
tk_root,
|
||||
TkTransportParams(
|
||||
audio_out_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
)
|
||||
|
||||
pipeline = Pipeline([transport.input(), MirrorProcessor(), tk_transport.output()])
|
||||
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
params=PipelineParams(),
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
async def run_tk():
|
||||
while not task.has_finished():
|
||||
tk_root.update()
|
||||
tk_root.update_idletasks()
|
||||
await asyncio.sleep(0.1)
|
||||
|
||||
@transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, client):
|
||||
logger.info(f"Client connected")
|
||||
await maybe_capture_participant_camera(transport, client, framerate=30)
|
||||
|
||||
@transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info(f"Client disconnected")
|
||||
await task.cancel()
|
||||
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
|
||||
await asyncio.gather(runner.run(task), run_tk())
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
108
examples/video-processing/mirror.py
Normal file
108
examples/video-processing/mirror.py
Normal file
@@ -0,0 +1,108 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
|
||||
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.frames.frames import (
|
||||
Frame,
|
||||
InputAudioRawFrame,
|
||||
InputImageRawFrame,
|
||||
OutputAudioRawFrame,
|
||||
OutputImageRawFrame,
|
||||
)
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineParams, PipelineTask
|
||||
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.transports.base_transport import BaseTransport, TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
|
||||
class MirrorProcessor(FrameProcessor):
|
||||
async def process_frame(self, frame: Frame, direction: FrameDirection):
|
||||
await super().process_frame(frame, direction)
|
||||
|
||||
if isinstance(frame, InputAudioRawFrame):
|
||||
await self.push_frame(
|
||||
OutputAudioRawFrame(
|
||||
audio=frame.audio,
|
||||
sample_rate=frame.sample_rate,
|
||||
num_channels=frame.num_channels,
|
||||
)
|
||||
)
|
||||
elif isinstance(frame, InputImageRawFrame):
|
||||
await self.push_frame(
|
||||
OutputImageRawFrame(image=frame.image, size=frame.size, format=frame.format)
|
||||
)
|
||||
else:
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
|
||||
# We use lambdas to defer transport parameter creation until the transport
|
||||
# type is selected at runtime.
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
video_out_width=1280,
|
||||
video_out_height=720,
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
async def run_bot(transport: BaseTransport, runner_args: RunnerArguments):
|
||||
logger.info(f"Starting bot")
|
||||
|
||||
pipeline = Pipeline([transport.input(), MirrorProcessor(), transport.output()])
|
||||
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
params=PipelineParams(),
|
||||
idle_timeout_secs=runner_args.pipeline_idle_timeout_secs,
|
||||
)
|
||||
|
||||
@transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, client):
|
||||
logger.info(f"Client connected")
|
||||
|
||||
@transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info(f"Client disconnected")
|
||||
await task.cancel()
|
||||
|
||||
runner = PipelineRunner(handle_sigint=runner_args.handle_sigint)
|
||||
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport, runner_args)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
173
examples/video-processing/video-processing.py
Normal file
173
examples/video-processing/video-processing.py
Normal file
@@ -0,0 +1,173 @@
|
||||
#
|
||||
# Copyright (c) 2024-2026, Daily
|
||||
#
|
||||
# SPDX-License-Identifier: BSD 2-Clause License
|
||||
#
|
||||
import os
|
||||
|
||||
import cv2
|
||||
import numpy as np
|
||||
from dotenv import load_dotenv
|
||||
from loguru import logger
|
||||
|
||||
from pipecat.audio.vad.silero import SileroVADAnalyzer
|
||||
from pipecat.frames.frames import Frame, InputImageRawFrame, LLMRunFrame, OutputImageRawFrame
|
||||
from pipecat.pipeline.pipeline import Pipeline
|
||||
from pipecat.pipeline.runner import PipelineRunner
|
||||
from pipecat.pipeline.task import PipelineParams, PipelineTask
|
||||
from pipecat.processors.aggregators.llm_context import LLMContext
|
||||
from pipecat.processors.aggregators.llm_response_universal import (
|
||||
LLMContextAggregatorPair,
|
||||
LLMUserAggregatorParams,
|
||||
)
|
||||
from pipecat.processors.frame_processor import FrameDirection, FrameProcessor
|
||||
from pipecat.runner.types import RunnerArguments
|
||||
from pipecat.runner.utils import create_transport
|
||||
from pipecat.services.google.gemini_live.llm import GeminiLiveLLMService
|
||||
from pipecat.transports.base_transport import TransportParams
|
||||
from pipecat.transports.daily.transport import DailyParams, DailyTransport
|
||||
|
||||
load_dotenv(override=True)
|
||||
|
||||
transport_params = {
|
||||
"daily": lambda: DailyParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
audio_out_10ms_chunks=2,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
),
|
||||
"webrtc": lambda: TransportParams(
|
||||
audio_in_enabled=True,
|
||||
audio_out_enabled=True,
|
||||
audio_out_10ms_chunks=2,
|
||||
video_in_enabled=True,
|
||||
video_out_enabled=True,
|
||||
video_out_is_live=True,
|
||||
),
|
||||
}
|
||||
|
||||
|
||||
class EdgeDetectionProcessor(FrameProcessor):
|
||||
def __init__(self, video_out_width, video_out_height: int):
|
||||
super().__init__()
|
||||
self._video_out_width = video_out_width
|
||||
self._video_out_height = video_out_height
|
||||
|
||||
async def process_frame(self, frame: Frame, direction: FrameDirection):
|
||||
await super().process_frame(frame, direction)
|
||||
|
||||
# Send back the user's camera video with edge detection applied
|
||||
if isinstance(frame, InputImageRawFrame) and frame.transport_source == "camera":
|
||||
# Convert bytes to NumPy array
|
||||
img = np.frombuffer(frame.image, dtype=np.uint8).reshape(
|
||||
(frame.size[1], frame.size[0], 3)
|
||||
)
|
||||
|
||||
# perform edge detection only on camera frames
|
||||
img = cv2.cvtColor(cv2.Canny(img, 100, 200), cv2.COLOR_GRAY2BGR)
|
||||
|
||||
# convert the size if needed
|
||||
desired_size = (self._video_out_width, self._video_out_height)
|
||||
if frame.size != desired_size:
|
||||
resized_image = cv2.resize(img, desired_size)
|
||||
out_frame = OutputImageRawFrame(resized_image.tobytes(), desired_size, frame.format)
|
||||
await self.push_frame(out_frame)
|
||||
else:
|
||||
out_frame = OutputImageRawFrame(
|
||||
image=img.tobytes(), size=frame.size, format=frame.format
|
||||
)
|
||||
await self.push_frame(out_frame)
|
||||
else:
|
||||
await self.push_frame(frame, direction)
|
||||
|
||||
|
||||
SYSTEM_INSTRUCTION = f"""
|
||||
"You are Gemini Chatbot, a friendly, helpful robot.
|
||||
|
||||
Your goal is to demonstrate your capabilities in a succinct way.
|
||||
|
||||
Your output will be spoken aloud, so avoid special characters that can't easily be spoken, such as emojis or bullet points.
|
||||
|
||||
Respond to what the user said in a creative and helpful way. Keep your responses brief. One or two sentences at most.
|
||||
"""
|
||||
|
||||
|
||||
async def run_bot(pipecat_transport):
|
||||
llm = GeminiLiveLLMService(
|
||||
api_key=os.getenv("GOOGLE_API_KEY"),
|
||||
voice_id="Puck", # Aoede, Charon, Fenrir, Kore, Puck
|
||||
transcribe_user_audio=True,
|
||||
system_instruction=SYSTEM_INSTRUCTION,
|
||||
)
|
||||
|
||||
messages = [
|
||||
{
|
||||
"role": "developer",
|
||||
"content": "Start by greeting the user warmly and introducing yourself.",
|
||||
}
|
||||
]
|
||||
|
||||
context = LLMContext(messages)
|
||||
user_aggregator, assistant_aggregator = LLMContextAggregatorPair(
|
||||
context,
|
||||
user_params=LLMUserAggregatorParams(vad_analyzer=SileroVADAnalyzer()),
|
||||
)
|
||||
|
||||
pipeline = Pipeline(
|
||||
[
|
||||
pipecat_transport.input(),
|
||||
user_aggregator,
|
||||
llm, # LLM
|
||||
EdgeDetectionProcessor(
|
||||
pipecat_transport._params.video_out_width,
|
||||
pipecat_transport._params.video_out_height,
|
||||
), # Sending the video back to the user
|
||||
pipecat_transport.output(),
|
||||
assistant_aggregator,
|
||||
]
|
||||
)
|
||||
|
||||
task = PipelineTask(
|
||||
pipeline,
|
||||
params=PipelineParams(
|
||||
enable_metrics=True,
|
||||
enable_usage_metrics=True,
|
||||
),
|
||||
)
|
||||
|
||||
@task.rtvi.event_handler("on_client_ready")
|
||||
async def on_client_ready(rtvi):
|
||||
logger.info("Pipecat client ready.")
|
||||
# Kick off the conversation.
|
||||
await task.queue_frames([LLMRunFrame()])
|
||||
|
||||
@pipecat_transport.event_handler("on_client_connected")
|
||||
async def on_client_connected(transport, participant):
|
||||
logger.info("Pipecat Client connected")
|
||||
if isinstance(transport, DailyTransport):
|
||||
await pipecat_transport.capture_participant_video(participant["id"], framerate=30)
|
||||
else:
|
||||
await pipecat_transport.capture_participant_video("camera")
|
||||
|
||||
@pipecat_transport.event_handler("on_client_disconnected")
|
||||
async def on_client_disconnected(transport, client):
|
||||
logger.info("Pipecat Client disconnected")
|
||||
await task.cancel()
|
||||
|
||||
runner = PipelineRunner(handle_sigint=False, force_gc=True)
|
||||
|
||||
await runner.run(task)
|
||||
|
||||
|
||||
async def bot(runner_args: RunnerArguments):
|
||||
"""Main bot entry point compatible with Pipecat Cloud."""
|
||||
transport = await create_transport(runner_args, transport_params)
|
||||
await run_bot(transport)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
from pipecat.runner.run import main
|
||||
|
||||
main()
|
||||
Reference in New Issue
Block a user