18411445. IMAGE RESTORATION FOR THROUGH-DISPLAY IMAGING simplified abstract (Microsoft Technology Licensing, LLC)

IMAGE RESTORATION FOR THROUGH-DISPLAY IMAGING

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

Yuqian Zhou of Urbana IL (US)

Timothy Andrew Large of Bellevue WA (US)

Se Hoon Lim of Bellevue WA (US)

Neil Emerton of Redmond WA (US)

Yonghuan David Ren of El Cerrito CA (US)

IMAGE RESTORATION FOR THROUGH-DISPLAY IMAGING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18411445 titled 'IMAGE RESTORATION FOR THROUGH-DISPLAY IMAGING

Simplified Explanation

The patent application relates to restoring degraded images acquired via a behind-display camera using machine learning models.

• Training method:

 - Input training image pairs into the machine learning model.
 - Each pair includes an undegraded image and a degraded image.
 - Train the model to generate missing frequency information from the degraded images.

---

1. 1. 1. Potential Applications of this Technology

- Image restoration in surveillance systems - Enhancing image quality in medical imaging

1. 1. 1. Problems Solved by this Technology

- Restoring degraded images for better analysis - Improving image quality for accurate identification

1. 1. 1. Benefits of this Technology

- Enhanced image quality for various applications - Efficient restoration process using machine learning

1. 1. 1. Potential Commercial Applications of this Technology
         1. Enhanced Image Restoration for Surveillance Systems

1. 1. 1. Possible Prior Art

No prior art is known at this time.

---

1. 1. Unanswered Questions

1. 1. 1. How does this technology handle different types of image degradation?

The patent application does not specify how the machine learning model adapts to various forms of image degradation.

1. 1. 1. What is the computational cost of training the machine learning model?

The patent application does not mention the computational resources required for training the model.

Original Abstract Submitted

Examples are disclosed that relate to the restoration of degraded images acquired via a behind-display camera. One example provides a method of training a machine learning model, the method comprising inputting training image pairs into the machine learning model, each training image pair comprising an undegraded image and a degraded image that represents an appearance of the undegraded image to a behind-display camera, and training the machine learning model using the training image pairs to generate frequency information that is missing from the degraded images.