Spaces:
Running
Running
File size: 3,254 Bytes
35d2acb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 |
import streamlit as st
from faster_whisper import WhisperModel
from transformers import pipeline
from textblob import TextBlob
import numpy as np
import queue
import threading
import time
# Initialize shared state
audio_queue = queue.Queue()
transcription_results = []
ai_detection_results = []
# Global stop event for threads
stop_event = threading.Event()
def initialize_model():
"""Initialize Whisper model and AI detectors."""
st.session_state.model = WhisperModel(model_size="small", device="cpu", compute_type="int8")
st.session_state.ai_detector = pipeline('text-classification', model='roberta-base-openai-detector')
def advanced_ai_detection(text, ai_detector):
"""Perform AI detection on the text."""
if len(text.split()) < 5: # Skip short texts
return "Insufficient Data"
result = ai_detector(text)[0]
return result
def analyze_text_linguistics(text):
"""Perform linguistic analysis on the text."""
blob = TextBlob(text)
return {
"sentiment": blob.sentiment.polarity,
"subjectivity": blob.sentiment.subjectivity
}
def transcribe_audio(audio_chunk, model):
"""Transcribe audio using the Whisper model."""
segments, _ = model.transcribe(audio_chunk, language="en")
return [segment.text for segment in segments]
def run_app():
"""Main Streamlit app function."""
st.title("AI Speech Detector")
st.subheader("Real-Time Speech Transcription and AI Detection")
st.text("This app transcribes audio input and detects if the text is AI-generated.")
# Sidebar for controls
st.sidebar.title("Controls")
start_button = st.sidebar.button("Start Recording")
stop_button = st.sidebar.button("Stop Recording")
if "model" not in st.session_state:
st.text("Loading AI models...")
initialize_model()
st.text("Models loaded successfully!")
# Display transcript
st.text_area("Real-Time Transcript", height=200, key="transcript")
# Display AI detection results
st.text_area("AI Detection Results", height=200, key="ai_detection")
if start_button:
st.session_state.is_recording = True
threading.Thread(target=process_audio_stream).start()
if stop_button:
st.session_state.is_recording = False
stop_event.set()
def process_audio_stream():
"""Simulated audio capture and processing."""
model = st.session_state.model
ai_detector = st.session_state.ai_detector
while not stop_event.is_set():
try:
# Simulate real-time audio input
fake_audio_chunk = np.random.rand(16000 * 3).astype(np.float32) # 3 seconds of fake audio
transcription = transcribe_audio(fake_audio_chunk, model)
# Update real-time transcript
for text in transcription:
st.session_state.transcript += text + "\n"
# Perform AI detection
ai_result = advanced_ai_detection(text, ai_detector)
st.session_state.ai_detection += f"Text: {text}\nResult: {ai_result}\n\n"
except Exception as e:
st.error(f"Error during transcription: {str(e)}")
break
stop_event.clear()
if __name__ == "__main__":
run_app()
|