Apify AI Image Upscaler & Face Restoration Actor

Elevate your visuals with our powerful Apify Actor, designed for superior AI image upscaling, intelligent image enhancement, and realistic face restoration. Transform low-resolution photos into stunning, high-quality masterpieces.

🌟 Key Features & Benefits

• Advanced AI Face Restoration: Utilizing cutting-edge neural networks, this actor meticulously restores and enhances faces, bringing unparalleled clarity and detail to portraits and old photos. Achieve flawless skin, sharp features, and natural expressions.
• High-Quality Image Upscaling: Increase image resolution up to 4x using advanced AI algorithms. Say goodbye to pixelation and blur; get crisp, clear, and vibrant images perfect for any use.
• Intelligent Image Enhancement: Beyond just upscaling, our AI enhances overall image quality, improving colors, contrast, and sharpness for a professional finish.
• Flexible Control (Fidelity Weight): Fine-tune the balance between preserving original identity and maximizing enhancement. Achieve your desired artistic effect while maintaining authenticity.
• Multiple Face Detection Models: Choose between highly accurate RetinaFace for precise detection in complex scenes or the lightning-fast YOLOv5 for rapid processing of simple portraits.
• Efficient Batch Processing: Enhance multiple images in a single run, saving time and streamlining your workflow for large datasets.
• Seamless Integration: Get enhanced images as public URLs, ready for direct use in your applications, websites, or databases.
• Robust & Reliable: Built for production, with comprehensive error handling and automatic model setup for a hassle-free experience.

💡 Ideal Use Cases

• Restore Old Photos: Breathe new life into vintage photographs, repairing damage, enhancing clarity, and bringing back cherished memories.
• Improve Low-Resolution Images: Upgrade blurry or pixelated images to high-definition quality for printing, web, or display.
• Enhance Portraits & Selfies: Achieve professional-grade enhancements for personal and professional photos, perfect for social media, portfolios, or family albums.
• E-commerce Product Images: Improve the visual appeal of product photos with enhanced details and clarity, boosting customer engagement.
• Digital Art & Design: Upscale and refine images for creative projects, ensuring high-quality output for illustrations, textures, and more.
• Forensic Image Enhancement: Improve clarity of images for analysis and identification.

🚀 How to Use

Simply provide an array of image URLs, and the actor will process them, returning enhanced versions.

Input Parameters

• imageUrls: Array of image URLs to process (supports JPG, PNG, BMP, TIFF).
• memory: Select the amount of RAM/memory to use (in GB). Options: 4 GB, 8 GB.

🔧 Configuration Tips

Optimize your results with these settings:

Fidelity Weight Guidelines

• 0.3-0.5: Maximum quality enhancement, may alter identity slightly.
• 0.5-0.7: Balanced enhancement and identity preservation (recommended for most cases).
• 0.7-1.0: Maximum identity preservation, moderate enhancement.

Detection Model Selection

• RetinaFace: More accurate face detection, ideal for complex scenes with multiple faces or challenging angles.
• YOLOv5: Faster processing, excellent for simple portraits and single-face images.

Upscale Factor Recommendations

• 1x: Enhancement only, no upscaling.
• 2x: Good balance of quality and speed.
• 3-4x: High resolution output, suitable for large prints or detailed viewing (may require more processing time).

❓ Troubleshooting & Tips

Common Issues

"No images were successfully processed"

• Ensure image URLs are accessible and valid.
• Verify image formats are supported (JPG, PNG, BMP, TIFF).

"No enhanced images were produced"

• The input images may not contain detectable faces.
• Try using a different detection model (RetinaFace vs. YOLOv5).
• Check if the onlyCenter parameter is too restrictive (if used).

Memory considerations

• For very large images or high upscale factors, consider processing fewer images per batch or using smaller input images if possible.

Performance Optimization

• For images with multiple faces, using onlyCenter=true can focus processing on the primary face.
• If you have pre-aligned face images, setting hasAlignedImages=true can speed up processing.
• Choosing the YOLOv5 detection model can lead to faster processing times for suitable images.