Create app.py

app.py

@@ -0,0 +1,74 @@
+import streamlit as st
+from PIL import Image
+from transformers import pipeline
+from gtts import gTTS
+import os
+
+# 加载 Hugging Face 模型
+image_to_text_model = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
+story_generator = pipeline("text-generation", model="facebook/opt-1.3b")
+
+# 图片 → 文字（生成描述）
+def img2text(image_path):
+    text = image_to_text_model(image_path)[0]["generated_text"]
+    return text
+
+# 文字 → 故事（生成完整故事）
+def text2story(text):
+    prompt = f"Write a fun and magical children's story based on this idea: {text}.\n\nOnce upon a time..."
+    story = story_generator(prompt, max_length=250, do_sample=True, temperature=0.8, top_p=0.9, repetition_penalty=1.2, truncation=True)[0]['generated_text']
+    return story
+
+# 故事 → 语音（TTS）
+def text2audio_gtts(story_text, filename="story.mp3"):
+    # 避免文件冲突
+    if os.path.exists(filename):
+        os.remove(filename)
+
+    # 限制 TTS 文本长度
+    max_chars = 500  # gTTS 可能不支持过长文本
+    story_text = story_text[:max_chars]
+
+    # 生成语音
+    tts = gTTS(text=story_text, lang="en")
+    tts.save(filename)
+
+    return filename
+
+# Streamlit Web UI
+st.set_page_config(page_title="AI Storyteller", page_icon="📖")
+st.header("📖 AI Storyteller: Turn Your Image into a Story with Audio")
+
+uploaded_file = st.file_uploader("Upload an Image...", type=["jpg", "png"])
+
+if uploaded_file:
+    # 保存图片到本地
+    image_path = "uploaded_image.jpg"
+    with open(image_path, "wb") as f:
+        f.write(uploaded_file.getbuffer())
+
+    # 读取并显示图片
+    image = Image.open(image_path)
+    st.image(image, caption="Uploaded Image", use_column_width=True)
+
+    # 生成图片描述
+    st.text("🔍 Generating image caption...")
+    caption = img2text(image_path)  # 这里改成文件路径
+    st.write("**Image Description:**", caption)
+
+    # 生成故事
+    st.text("📝 Generating story...")
+    story = text2story(caption)
+    st.write("**Generated Story:**")
+    st.write(story)
+
+    # 生成语音
+    st.text("🔊 Generating audio...")
+    audio_file = text2audio_gtts(story)
+
+    # 播放音频
+    st.audio(audio_file, format="audio/mp3")
+
+    # 下载按钮
+    with open(audio_file, "rb") as file:
+        st.download_button("📥 Download Audio", file, file_name="story.mp3")