18179855. INPAINTING DIGITAL IMAGES USING A HYBRID WIRE REMOVAL PIPELINE simplified abstract (Adobe Inc.)
INPAINTING DIGITAL IMAGES USING A HYBRID WIRE REMOVAL PIPELINE
Organization Name
Inventor(s)
Connelly Barnes of Seattle WA (US)
Elya Shechtman of Seattle WA (US)
Sohrab Amirghodsi of Seattle WA (US)
Xiaoyang Liu of Bellevue WA (US)
INPAINTING DIGITAL IMAGES USING A HYBRID WIRE REMOVAL PIPELINE - A simplified explanation of the abstract
This abstract first appeared for US patent application 18179855 titled 'INPAINTING DIGITAL IMAGES USING A HYBRID WIRE REMOVAL PIPELINE
Simplified Explanation: The patent application describes a system for inpainting digital images by removing wires using a hybrid wire removal pipeline that integrates multiple machine learning models.
- The system generates a wire segmentation from the digital image.
- It identifies specific wires or portions of wires in the segmentation.
- The system inpaints pixels corresponding to the wires using patch-based and deep inpainting models.
Key Features and Innovation:
- Hybrid wire removal pipeline integrating machine learning models.
- Wire segmentation generation and identification of specific wires.
- Inpainting of wire pixels using patch-based and deep inpainting models.
Potential Applications:
- Image editing software.
- Forensic analysis tools.
- Restoration of damaged photographs.
Problems Solved:
- Efficient removal of wires from digital images.
- Accurate identification of wire segments.
- Seamless inpainting of wire pixels.
Benefits:
- Improved image quality.
- Time-saving in image editing processes.
- Enhanced forensic analysis capabilities.
Commercial Applications: Wire removal technology for image editing software with applications in photography, forensics, and restoration services.
Questions about Wire Removal Technology: 1. How does the hybrid wire removal pipeline improve the inpainting process? 2. What are the main advantages of using machine learning models for wire removal in digital images?
Original Abstract Submitted
The present disclosure relates to systems, methods, and non-transitory computer readable media for inpainting a digital image using a hybrid wire removal pipeline. For example, the disclosed systems use a hybrid wire removal pipeline that integrates multiple machine learning models, such as a wire segmentation model, a hole separation model, a mask dilation model, a patch-based inpainting model, and a deep inpainting model. Using the hybrid wire removal pipeline, in some embodiments, the disclosed systems generate a wire segmentation from a digital image depicting one or more wires. The disclosed systems also utilize the hybrid wire removal pipeline to extract or identify portions of the wire segmentation that indicate specific wires or portions of wires. In certain embodiments, the disclosed systems further inpaint pixels of the digital image corresponding to the wires indicated by the wire segmentation mask using the patch-based inpainting model and/or the deep inpainting model.
