Spaces:
Paused
Paused
Download app.py from Ribakoo/Image-Enhancer: direct link, hf CLI and curl.
- Browser
- Download file 19 kB
-
https://huggingface.co/spaces/Ribakoo/Image-Enhancer/resolve/main/app.py
- Command line
-
hf download hf://spaces/Ribakoo/Image-Enhancer/app.py
-
curl -L -o app.py https://huggingface.co/spaces/Ribakoo/Image-Enhancer/resolve/main/app.py
19 kB
| """ | |
| VectorForge ⚡ - Pro Max UI/UX Edition | |
| AI 2D Character Super-Resolution & Mathematical Vectorization Engine | |
| Designed with UI/UX Pro Max Intelligence: Tech Startup Typography, Dark Glassmorphism, Micro-Interactions | |
| """ | |
| import os | |
| import base64 | |
| from pathlib import Path | |
| import urllib.request | |
| import time | |
| from PIL import Image | |
| import gradio as gr | |
| # Ensure models exist | |
| BASE_DIR = Path(__file__).resolve().parent | |
| MODELS_DIR = BASE_DIR / "models" | |
| MODELS_DIR.mkdir(parents=True, exist_ok=True) | |
| MODEL_URLS = { | |
| MODELS_DIR / "waifu2x_art_scale2x.onnx": "https://huggingface.co/deepghs/waifu2x_onnx/resolve/main/20230131/onnx_models/cunet/art/scale2x.onnx", | |
| MODELS_DIR / "waifu2x_art_noise1_scale2x.onnx": "https://huggingface.co/deepghs/waifu2x_onnx/resolve/main/20230131/onnx_models/cunet/art/noise1_scale2x.onnx" | |
| } | |
| for path, url in MODEL_URLS.items(): | |
| if not path.exists(): | |
| print(f"Downloading {path.name}...") | |
| urllib.request.urlretrieve(url, path) | |
| from engine.upscaler import AI2DUpscaler | |
| from engine.vectorizer import Vectorizer | |
| upscaler = AI2DUpscaler(models_dir=MODELS_DIR) | |
| STORAGE_DIR = BASE_DIR / "storage" | |
| STORAGE_DIR.mkdir(exist_ok=True) | |
| # Safe ZeroGPU decorator support | |
| try: | |
| import spaces | |
| gpu_decorator = spaces.GPU | |
| except (ImportError, Exception): | |
| def gpu_decorator(fn): | |
| return fn | |
| def image_to_base64_data_uri(img: Image.Image) -> str: | |
| import io | |
| buffered = io.BytesIO() | |
| img.save(buffered, format="PNG") | |
| b64 = base64.b64encode(buffered.getvalue()).decode("utf-8") | |
| return f"data:image/png;base64,{b64}" | |
| def process_pipeline(input_img, mode, preset, scale, denoise): | |
| if input_img is None: | |
| empty_html = """ | |
| <div class="empty-placeholder"> | |
| <div class="placeholder-icon">🎨</div> | |
| <h3>No Image Uploaded</h3> | |
| <p>Select an example below or upload a 2D character/icon to begin.</p> | |
| </div> | |
| """ | |
| return None, None, None, empty_html, "", "Upload an image to start." | |
| t0 = time.time() | |
| timestamp = int(time.time() * 1000) | |
| # Save input temp | |
| temp_in = STORAGE_DIR / f"temp_{timestamp}.png" | |
| if isinstance(input_img, Image.Image): | |
| pil_in = input_img.convert("RGBA") | |
| else: | |
| pil_in = Image.fromarray(input_img).convert("RGBA") | |
| pil_in.save(temp_in, format="PNG") | |
| orig_w, orig_h = pil_in.size | |
| in_b64 = image_to_base64_data_uri(pil_in) | |
| out_png_path = None | |
| out_svg_path = None | |
| enhanced_pil = None | |
| preset_map = { | |
| "2D Character / Anime": "character", | |
| "Flat Color / Icon": "flat_icon", | |
| "Detailed Lineart": "detailed_lineart", | |
| "Monochrome / Ink": "monochrome", | |
| } | |
| v_preset = preset_map.get(preset, "character") | |
| scale_factor = int(scale.replace("x", "")) | |
| # Step 1: AI Super-Resolution (if selected) | |
| if "AI" in mode or "Hybrid" in mode: | |
| out_png_path = STORAGE_DIR / f"enhanced_{timestamp}.png" | |
| upscaler.upscale_file(temp_in, out_png_path, scale=scale_factor, denoise=denoise) | |
| enhanced_pil = Image.open(out_png_path) | |
| enhanced_w, enhanced_h = enhanced_pil.size | |
| else: | |
| enhanced_w, enhanced_h = orig_w, orig_h | |
| # Step 2: Vectorization (if selected) | |
| svg_code = "" | |
| if "Vector" in mode or "Hybrid" in mode: | |
| out_svg_path = STORAGE_DIR / f"vector_{timestamp}.svg" | |
| source_for_vector = out_png_path if (out_png_path and "Hybrid" in mode) else temp_in | |
| Vectorizer.vectorize_file(source_for_vector, out_svg_path, preset=v_preset) | |
| svg_code = out_svg_path.read_text(encoding="utf-8") | |
| elapsed = round(time.time() - t0, 2) | |
| after_b64 = image_to_base64_data_uri(enhanced_pil if enhanced_pil else pil_in) | |
| # UI/UX Pro Max Interactive Comparison Slider HTML Component | |
| slider_html = f""" | |
| <div class="pro-slider-wrapper"> | |
| <div class="pro-slider-container" id="slider-box" onmousemove="moveCompare(event)" ontouchmove="moveCompareTouch(event)"> | |
| <img src="{after_b64}" class="slider-img-after" alt="Enhanced Result"> | |
| <div class="slider-before-mask" id="slider-mask"> | |
| <img src="{in_b64}" class="slider-img-before" id="before-img-el" alt="Original Low-Res"> | |
| </div> | |
| <div class="slider-divider" id="slider-line"> | |
| <div class="slider-pill"> | |
| <span>❮ ❯</span> | |
| </div> | |
| </div> | |
| <span class="badge-tag tag-left">ORIGINAL ({orig_w}×{orig_h})</span> | |
| <span class="badge-tag tag-right">ENHANCED 4× ({enhanced_w}×{enhanced_h})</span> | |
| </div> | |
| </div> | |
| <script> | |
| function moveCompare(e) {{ | |
| var box = document.getElementById('slider-box'); | |
| if(!box) return; | |
| var rect = box.getBoundingClientRect(); | |
| var x = e.clientX - rect.left; | |
| var pct = Math.max(0, Math.min(100, (x / rect.width) * 100)); | |
| updateSlider(pct); | |
| }} | |
| function moveCompareTouch(e) {{ | |
| var box = document.getElementById('slider-box'); | |
| if(!box || !e.touches.length) return; | |
| var rect = box.getBoundingClientRect(); | |
| var x = e.touches[0].clientX - rect.left; | |
| var pct = Math.max(0, Math.min(100, (x / rect.width) * 100)); | |
| updateSlider(pct); | |
| }} | |
| function updateSlider(pct) {{ | |
| var mask = document.getElementById('slider-mask'); | |
| var line = document.getElementById('slider-line'); | |
| var box = document.getElementById('slider-box'); | |
| var beforeImg = document.getElementById('before-img-el'); | |
| if(mask && line && box) {{ | |
| mask.style.width = pct + '%'; | |
| line.style.left = pct + '%'; | |
| if(beforeImg) beforeImg.style.width = box.offsetWidth + 'px'; | |
| }} | |
| }} | |
| setTimeout(function() {{ updateSlider(50); }}, 300); | |
| </script> | |
| """ | |
| # UI/UX Pro Max Metric Cards HTML | |
| svg_size_kb = round(out_svg_path.stat().st_size / 1024, 1) if out_svg_path else 0 | |
| png_size_kb = round(out_png_path.stat().st_size / 1024, 1) if out_png_path else 0 | |
| pixel_increase = round(((enhanced_w * enhanced_h) / (orig_w * orig_h) - 1) * 100) if (orig_w * orig_h) else 0 | |
| metrics_html = f""" | |
| <div class="metrics-grid"> | |
| <div class="metric-card"> | |
| <span class="metric-label">Resolution Boost</span> | |
| <span class="metric-value accent-indigo">{orig_w}×{orig_h} → {enhanced_w}×{enhanced_h}</span> | |
| <span class="metric-sub">+{pixel_increase}% total pixel clarity</span> | |
| </div> | |
| <div class="metric-card"> | |
| <span class="metric-label">Inference Latency</span> | |
| <span class="metric-value accent-emerald">{elapsed}s</span> | |
| <span class="metric-sub">ZeroGPU & CUNet acceleration</span> | |
| </div> | |
| <div class="metric-card"> | |
| <span class="metric-label">Scalable Vector</span> | |
| <span class="metric-value accent-pink">{svg_size_kb} KB SVG</span> | |
| <span class="metric-sub">Mathematical Bezier splines</span> | |
| </div> | |
| <div class="metric-card"> | |
| <span class="metric-label">Raster Master</span> | |
| <span class="metric-value">{png_size_kb} KB PNG</span> | |
| <span class="metric-sub">Artifacts & noise removed</span> | |
| </div> | |
| </div> | |
| """ | |
| return ( | |
| enhanced_pil if enhanced_pil else pil_in, | |
| str(out_png_path) if out_png_path else None, | |
| str(out_svg_path) if out_svg_path else None, | |
| slider_html, | |
| svg_code, | |
| metrics_html | |
| ) | |
| # UI/UX Pro Max Custom CSS & Theme Styling | |
| CUSTOM_CSS = """ | |
| @import url('https://fonts.googleapis.com/css2?family=DM+Sans:wght@400;500;700&family=JetBrains+Mono:wght@400;500;700&family=Space+Grotesk:wght@500;600;700&display=swap'); | |
| :root { | |
| --bg-obsidian: #090d16; | |
| --card-surface: rgba(17, 24, 39, 0.75); | |
| --border-glass: rgba(255, 255, 255, 0.08); | |
| --accent-indigo: #6366f1; | |
| --accent-emerald: #10b981; | |
| --accent-pink: #ec4899; | |
| } | |
| body, .gradio-container { | |
| background-color: var(--bg-obsidian) !important; | |
| font-family: 'DM Sans', -apple-system, sans-serif !important; | |
| color: #f1f5f9 !important; | |
| } | |
| h1, h2, h3, .pro-title { | |
| font-family: 'Space Grotesk', sans-serif !important; | |
| letter-spacing: -0.02em !important; | |
| } | |
| /* Glassmorphism Cards */ | |
| .glass-panel { | |
| background: var(--card-surface) !important; | |
| backdrop-filter: blur(16px) !important; | |
| -webkit-backdrop-filter: blur(16px) !important; | |
| border: 1px solid var(--border-glass) !important; | |
| border-radius: 16px !important; | |
| padding: 20px !important; | |
| box-shadow: 0 10px 30px -10px rgba(0, 0, 0, 0.5) !important; | |
| } | |
| /* Hero Badge */ | |
| .hero-badge { | |
| display: inline-flex; | |
| align-items: center; | |
| gap: 6px; | |
| padding: 4px 12px; | |
| border-radius: 9999px; | |
| background: rgba(99, 102, 241, 0.12); | |
| border: 1px solid rgba(99, 102, 241, 0.3); | |
| color: #a5b4fc; | |
| font-size: 12px; | |
| font-weight: 600; | |
| margin-bottom: 8px; | |
| } | |
| /* Interactive Split Comparison Slider */ | |
| .pro-slider-wrapper { | |
| width: 100%; | |
| margin: 8px 0; | |
| } | |
| .pro-slider-container { | |
| position: relative; | |
| width: 100%; | |
| height: 420px; | |
| border-radius: 14px; | |
| overflow: hidden; | |
| cursor: ew-resize; | |
| user-select: none; | |
| border: 1px solid rgba(255, 255, 255, 0.12); | |
| background-color: #0f172a; | |
| background-image: | |
| linear-gradient(45deg, #1e293b 25%, transparent 25%), | |
| linear-gradient(-45deg, #1e293b 25%, transparent 25%), | |
| linear-gradient(45deg, transparent 75%, #1e293b 75%), | |
| linear-gradient(-45deg, transparent 75%, #1e293b 75%); | |
| background-size: 16px 16px; | |
| background-position: 0 0, 0 8px, 8px -8px, -8px 0px; | |
| } | |
| .slider-img-after { | |
| position: absolute; | |
| top: 0; | |
| left: 0; | |
| width: 100%; | |
| height: 100%; | |
| object-fit: contain; | |
| pointer-events: none; | |
| } | |
| .slider-before-mask { | |
| position: absolute; | |
| top: 0; | |
| left: 0; | |
| width: 50%; | |
| height: 100%; | |
| overflow: hidden; | |
| border-right: 2px solid var(--accent-indigo); | |
| pointer-events: none; | |
| } | |
| .slider-img-before { | |
| position: absolute; | |
| top: 0; | |
| left: 0; | |
| height: 100%; | |
| object-fit: contain; | |
| pointer-events: none; | |
| } | |
| .slider-divider { | |
| position: absolute; | |
| top: 0; | |
| bottom: 0; | |
| left: 50%; | |
| width: 2px; | |
| background: var(--accent-indigo); | |
| pointer-events: none; | |
| box-shadow: 0 0 12px rgba(99, 102, 241, 0.8); | |
| } | |
| .slider-pill { | |
| position: absolute; | |
| top: 50%; | |
| left: 50%; | |
| transform: translate(-50%, -50%); | |
| width: 38px; | |
| height: 38px; | |
| border-radius: 50%; | |
| background: #4f46e5; | |
| color: white; | |
| display: flex; | |
| align-items: center; | |
| justify-content: center; | |
| font-size: 12px; | |
| font-weight: bold; | |
| box-shadow: 0 4px 16px rgba(79, 70, 229, 0.5); | |
| } | |
| .badge-tag { | |
| position: absolute; | |
| bottom: 12px; | |
| font-family: 'JetBrains Mono', monospace; | |
| font-size: 10px; | |
| font-weight: 600; | |
| padding: 4px 8px; | |
| border-radius: 6px; | |
| backdrop-filter: blur(8px); | |
| pointer-events: none; | |
| z-index: 5; | |
| } | |
| .tag-left { | |
| left: 12px; | |
| background: rgba(15, 23, 42, 0.85); | |
| border: 1px solid rgba(255, 255, 255, 0.15); | |
| color: #94a3b8; | |
| } | |
| .tag-right { | |
| right: 12px; | |
| background: rgba(49, 46, 129, 0.85); | |
| border: 1px solid rgba(129, 140, 248, 0.3); | |
| color: #c7d2fe; | |
| } | |
| /* Empty Placeholder */ | |
| .empty-placeholder { | |
| height: 380px; | |
| display: flex; | |
| flex-direction: column; | |
| align-items: center; | |
| justify-content: center; | |
| text-align: center; | |
| border: 2px dashed rgba(255, 255, 255, 0.1); | |
| border-radius: 14px; | |
| color: #64748b; | |
| } | |
| .placeholder-icon { | |
| font-size: 40px; | |
| margin-bottom: 12px; | |
| } | |
| /* Metrics Grid */ | |
| .metrics-grid { | |
| display: grid; | |
| grid-template-columns: repeat(auto-fit, minmax(130px, 1fr)); | |
| gap: 10px; | |
| margin-top: 14px; | |
| } | |
| .metric-card { | |
| background: rgba(15, 23, 42, 0.6); | |
| border: 1px solid rgba(255, 255, 255, 0.06); | |
| border-radius: 10px; | |
| padding: 10px 14px; | |
| display: flex; | |
| flex-direction: column; | |
| } | |
| .metric-label { | |
| font-size: 10px; | |
| text-transform: uppercase; | |
| font-weight: 700; | |
| color: #64748b; | |
| letter-spacing: 0.05em; | |
| } | |
| .metric-value { | |
| font-family: 'JetBrains Mono', monospace; | |
| font-size: 14px; | |
| font-weight: 700; | |
| color: #f8fafc; | |
| margin: 3px 0; | |
| } | |
| .metric-sub { | |
| font-size: 10px; | |
| color: #94a3b8; | |
| } | |
| .accent-indigo { color: #818cf8 !important; } | |
| .accent-emerald { color: #34d399 !important; } | |
| .accent-pink { color: #f472b6 !important; } | |
| /* Enhanced Button */ | |
| .primary-btn { | |
| background: linear-gradient(135deg, #4f46e5 0%, #6366f1 100%) !important; | |
| border: none !important; | |
| color: white !important; | |
| font-weight: 600 !important; | |
| box-shadow: 0 4px 14px rgba(79, 70, 229, 0.4) !important; | |
| transition: all 0.2s ease !important; | |
| } | |
| .primary-btn:hover { | |
| box-shadow: 0 6px 20px rgba(79, 70, 229, 0.6) !important; | |
| transform: translateY(-1px) !important; | |
| } | |
| """ | |
| with gr.Blocks( | |
| title="VectorForge ⚡ | AI 2D Character Super-Resolution & Vectorization", | |
| css=CUSTOM_CSS, | |
| theme=gr.themes.Base( | |
| primary_hue="indigo", | |
| neutral_hue="slate", | |
| font=[gr.themes.GoogleFont("DM Sans"), "sans-serif"] | |
| ) | |
| ) as demo: | |
| with gr.Row(): | |
| gr.HTML( | |
| """ | |
| <div style="text-align: center; max-width: 800px; margin: 0 auto; padding: 15px 0 10px;"> | |
| <div class="hero-badge">⚡ ZeroGPU Powered · Rust Vector Core</div> | |
| <h1 style="font-size: 2.2rem; font-weight: 800; margin: 6px 0; background: linear-gradient(to right, #ffffff, #c7d2fe, #818cf8); -webkit-background-clip: text; -webkit-text-fill-color: transparent;"> | |
| VectorForge | |
| </h1> | |
| <p style="color: #94a3b8; font-size: 0.95rem; margin-top: 4px;"> | |
| Transform low-resolution 2D character art, stickers, and icons into crisp 4× PNGs and infinitely-scalable mathematical SVG vectors. | |
| </p> | |
| </div> | |
| """ | |
| ) | |
| with gr.Row(): | |
| # Left: Controls & Upload | |
| with gr.Column(scale=5, elem_classes=["glass-panel"]): | |
| gr.Markdown("### 📥 Source Character Artwork") | |
| input_image = gr.Image( | |
| type="pil", | |
| label="Drop PNG / JPG with transparent or solid background", | |
| elem_id="source-image-input" | |
| ) | |
| # Interactive 1-click preset examples | |
| sample_avatar = str(BASE_DIR / "examples" / "mascot_avatar.png") | |
| sample_badge = str(BASE_DIR / "examples" / "game_badge.png") | |
| gr.Markdown("#### ⚡ Quick Try (1-Click Sample)") | |
| gr.Examples( | |
| examples=[ | |
| [sample_avatar, "Hybrid (AI Clean + SVG Vector)", "2D Character / Anime", "4x", True], | |
| [sample_badge, "Hybrid (AI Clean + SVG Vector)", "Flat Color / Icon", "4x", True], | |
| ], | |
| inputs=[input_image], | |
| label="Load Sample Characters" | |
| ) | |
| gr.Markdown("### ⚙️ Enhancement Controls") | |
| mode_radio = gr.Radio( | |
| choices=[ | |
| "Hybrid (AI Clean + SVG Vector)", | |
| "AI 4x (Crisp PNG)", | |
| "Direct Vectorize (SVG)" | |
| ], | |
| value="Hybrid (AI Clean + SVG Vector)", | |
| label="Processing Mode", | |
| info="Hybrid cleans pixel noise with neural CUNet then fits mathematical SVG bezier curves" | |
| ) | |
| with gr.Row(): | |
| preset_dropdown = gr.Dropdown( | |
| choices=["2D Character / Anime", "Flat Color / Icon", "Detailed Lineart", "Monochrome / Ink"], | |
| value="2D Character / Anime", | |
| label="Art Preset" | |
| ) | |
| scale_dropdown = gr.Dropdown( | |
| choices=["4x", "2x"], | |
| value="4x", | |
| label="Scale Multiplier" | |
| ) | |
| denoise_check = gr.Checkbox( | |
| value=True, | |
| label="Denoise Compression Artifacts", | |
| info="Cleans JPEG ringing and rough line edges" | |
| ) | |
| process_btn = gr.Button( | |
| "✨ Enhance & Vectorize", | |
| variant="primary", | |
| size="lg", | |
| elem_classes=["primary-btn"] | |
| ) | |
| # Right: Interactive Results & Telemetry | |
| with gr.Column(scale=7, elem_classes=["glass-panel"]): | |
| gr.Markdown("### 🔬 Live Inspection & Exports") | |
| # Real-time Metrics Card Bar | |
| metrics_display = gr.HTML( | |
| """ | |
| <div class="metrics-grid"> | |
| <div class="metric-card"><span class="metric-label">Engine</span><span class="metric-value accent-indigo">Ready</span></div> | |
| <div class="metric-card"><span class="metric-label">Hardware</span><span class="metric-value accent-emerald">ZeroGPU</span></div> | |
| <div class="metric-card"><span class="metric-label">Mode</span><span class="metric-value accent-pink">Hybrid</span></div> | |
| </div> | |
| """ | |
| ) | |
| # Output Tabs | |
| with gr.Tabs(): | |
| with gr.TabItem("↔️ Split Comparison Slider"): | |
| slider_output = gr.HTML( | |
| """ | |
| <div class="empty-placeholder"> | |
| <div class="placeholder-icon">🎨</div> | |
| <h3>Awaiting Image</h3> | |
| <p>Upload a 2D character or click an example on the left to see the split slider comparison.</p> | |
| </div> | |
| """ | |
| ) | |
| with gr.TabItem("🖼️ High-Res Preview"): | |
| preview_output = gr.Image(type="pil", label="High-Resolution 4x Output", interactive=False) | |
| with gr.TabItem("📄 SVG Code Inspector"): | |
| svg_code_output = gr.Code(label="Scalable Vector Graphic (SVG XML Code)", language="html") | |
| # One-Click Downloads | |
| with gr.Row(): | |
| download_svg = gr.File(label="Download Scalable Vector (.SVG)") | |
| download_png = gr.File(label="Download High-Res (.PNG)") | |
| # Event binding | |
| process_btn.click( | |
| fn=process_pipeline, | |
| inputs=[input_image, mode_radio, preset_dropdown, scale_dropdown, denoise_check], | |
| outputs=[preview_output, download_png, download_svg, slider_output, svg_code_output, metrics_display] | |
| ) | |
| if __name__ == "__main__": | |
| demo.launch(server_name="0.0.0.0", server_port=7860) | |