README.md

---
title: AI Creative Production Suite
emoji: 🎬
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: "5.0.0"
app_file: app.py
pinned: false
---

# 🎬 AI Creative Production Suite

A modular, all-in-one toolkit for AI-powered creative content production — built entirely with open-source models from Hugging Face.

**Live Space:** [https://huggingface.co/spaces/bleckhert/creative-suite](https://huggingface.co/spaces/bleckhert/creative-suite)

---

## 🛠️ Tools Included

| Tool | Description | Models Used |
|------|-------------|-------------|
| **🎨 Image Generator** | Text-to-image generation with SDXL | `stabilityai/stable-diffusion-xl-base-1.0` |
| **✨ Image Enhancer** | Upscale + skin texture smoothing | OpenCV bilateral filter + sharpening |
| **🧍 Pose Editor** | Multi-angle view generation from character reference | Pose extraction + image processing |
| **🔄 Image Variations** | Style variations from reference (cinematic, oil painting, neon, etc.) | Image filters + compositing |
| **🎬 Video Generator** | Text-to-video clips | `THUDM/CogVideoX-2b` (with fallback) |
| **🎵 Audio Generator** | Music + sound effects generation | `facebook/musicgen-*`, `facebook/audiogen-medium` |
| **📖 Plot Generator** | Scene-by-scene plot outlines with visual/audio prompts | Template + randomization engine |
| **🎞️ Film Editor** | Compose scenes + audio into final film | imageio + MoviePy pipeline |
| **💡 Prompt Helper** | Enhance prompts with style, lighting, camera angle tags | — |

---

## 🔒 Safety Controls

- **Content moderation** on all text prompts (blocks keywords for harmful/non-consensual content)
- **Visible watermarks** on all generated images/videos: "AI GENERATED - Fictional Content"
- **Invisible watermarking** via DCT-based encoding (attempted with `invisible-watermark`)
- **No face-swapping** tools — intentionally excluded
- **No voice cloning** — only instrumental music and environmental sound effects
- **Logging** of all prompt hashes for audit trail

---

## 🚀 How to Use

1. Go to the **Image Generator** tab → type a prompt → generate your first image
2. Use the **Prompt Helper** tab to craft better prompts with style/lighting tags
3. Feed generated images into **Image Enhancer** for upscaling
4. Use **Pose Editor** to get multiple camera angles of the same character
5. Generate **Variations** in different art styles
6. Use **Plot Generator** to create scene-by-scene scripts with visual/audio prompts
7. Generate **music** and **sound effects** in the Audio tab
8. Use **Video Generator** for motion clips
9. Assemble everything in the **Film Editor** → render your final production

---

## 📦 Dependencies

- gradio >= 5.0.0
- torch >= 2.0.0
- diffusers >= 0.32.0
- transformers >= 4.40.0
- accelerate >= 0.30.0
- opencv-python-headless
- imageio, imageio-ffmpeg
- pillow, numpy, scipy
- librosa, soundfile
- invisible-watermark

---

**Made with ❤️ using Hugging Face Diffusers, Transformers, and Gradio.**