Spaces:
Sleeping
Sleeping
File size: 4,273 Bytes
f5d12b4 |
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 |
# streamlit_ui.py
import streamlit as st
import requests
import base64
import tempfile
from backend.subtitle_utils import generate_srt_from_text, enhance_video_with_subtitles_and_bgm
st.set_page_config(
page_title="Prompta - Text to Media Generator",
page_icon="🎙️",
layout="wide",
initial_sidebar_state="expanded"
)
st.title("🎙️🖼️🎞️ Prompta - Text to Media Generator")
API_BASE = "http://localhost:8000"
def render_media(file_bytes, media_type, label):
b64 = base64.b64encode(file_bytes).decode()
if media_type == "audio":
st.audio(f"data:audio/wav;base64,{b64}", format="audio/wav")
elif media_type == "video":
st.video(f"data:video/mp4;base64,{b64}")
elif media_type == "image":
st.image(file_bytes, caption=label, use_column_width=True)
st.sidebar.header("🛠️ Settings")
TOKEN = st.sidebar.text_input("🔑 API Token", type="password")
HEADERS = {"Authorization": f"Bearer {TOKEN}"} if TOKEN else {}
tab = st.sidebar.radio("Select Task", ["Text to Audio", "Text to Image", "Text to Video"])
if tab == "Text to Audio":
st.subheader("🎤 Text to Audio")
text = st.text_area("Enter text")
voice = st.selectbox("Choose voice/language", ["en-US", "hi-IN", "te-IN", "ta-IN"])
if st.button("🔊 Generate Audio"):
with st.spinner("Generating audio..."):
r = requests.post(f"{API_BASE}/audio/generate", json={"text": text, "voice": voice}, headers=HEADERS)
if r.status_code == 200:
render_media(r.content, "audio", "Generated Audio")
else:
st.error(f"❌ Failed: {r.json().get('detail')}")
elif tab == "Text to Image":
st.subheader("🖼️ Text to Image")
prompt = st.text_area("Enter image prompt")
model = st.selectbox("Choose model", ["sdxl", "deepfloyd", "kandinsky"])
if st.button("🧠 Generate Image"):
with st.spinner("Generating image..."):
r = requests.post(f"{API_BASE}/image/generate", json={"prompt": prompt, "model": model}, headers=HEADERS)
if r.status_code == 200:
render_media(r.content, "image", "Generated Image")
else:
st.error(f"❌ Failed: {r.json().get('detail')}")
elif tab == "Text to Video":
st.subheader("🎞️ Text to Video")
prompt = st.text_area("Enter video prompt")
tone = st.selectbox("Tone", ["formal", "casual", "emotional", "documentary"])
domain = st.selectbox("Domain", ["health", "education", "governance", "entertainment"])
environment = st.selectbox("Environment", ["urban", "rural", "nature", "futuristic"])
transcript = st.text_area("Transcript (optional - for subtitles)", height=100)
enhance = st.checkbox("✨ Add Subtitles and Background Music")
if st.button("🎬 Generate Video"):
with st.spinner("Generating video..."):
r = requests.post(
f"{API_BASE}/video/generate",
json={"prompt": prompt, "tone": tone, "domain": domain, "environment": environment},
headers=HEADERS
)
if r.status_code == 200:
video_bytes = r.content
if enhance and transcript:
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp4") as tmp_vid:
tmp_vid.write(video_bytes)
tmp_vid_path = tmp_vid.name
srt_path = generate_srt_from_text(transcript, output_path="streamlit_subs.srt")
enhanced_path = "streamlit_final_video.mp4"
enhance_video_with_subtitles_and_bgm(
video_path=tmp_vid_path,
srt_path=srt_path,
bgm_path="default_bgm.mp3",
output_path=enhanced_path
)
with open(enhanced_path, "rb") as f:
render_media(f.read(), "video", "Enhanced Video")
else:
render_media(video_bytes, "video", "Generated Video")
else:
st.error(f"❌ Failed: {r.json().get('detail')}")
st.sidebar.markdown("---")
st.sidebar.info("Built with ❤️ for AI GovTech Challenge 2025")
|