18096972. METHOD AND DEVICE FOR REPRESENTING RENDERED SCENES simplified abstract (SAMSUNG ELECTRONICS CO., LTD.)
Contents
METHOD AND DEVICE FOR REPRESENTING RENDERED SCENES
Organization Name
Inventor(s)
Seokhwan Jang of Suwon-si (KR)
Donghoon Sagong of Suwon-si (KR)
METHOD AND DEVICE FOR REPRESENTING RENDERED SCENES - A simplified explanation of the abstract
This abstract first appeared for US patent application 18096972 titled 'METHOD AND DEVICE FOR REPRESENTING RENDERED SCENES
Simplified Explanation
The patent application describes a method and device for representing rendered scenes using a neural network model.
- Obtaining spatial information of sampling data
- Obtaining volume-rendering parameters from the neural network model using the spatial information
- Calculating a regularization term based on the distribution of the volume-rendering parameters
- Performing volume rendering based on the obtained parameters
- Training the neural network model to minimize a loss function based on the regularization term and the difference between a ground truth image and an estimated image from the volume rendering
Potential applications of this technology:
- Computer graphics
- Virtual reality
- Medical imaging
- Scientific visualization
Problems solved by this technology:
- Enhancing the quality of rendered scenes
- Improving the efficiency of volume rendering
- Training neural network models for better image estimation
Benefits of this technology:
- More realistic and accurate rendered scenes
- Faster rendering process
- Enhanced training of neural network models for image estimation
Original Abstract Submitted
Disclosed are a method and device for representing rendered scenes. A data processing method of training a neural network model includes obtaining spatial information of sampling data, obtaining one or more volume-rendering parameters by inputting the spatial information of the sampling data to the neural network model, obtaining a regularization term based on a distribution of the volume-rendering parameters, performing volume rendering based on the volume-rendering parameters, and training the neural network model to minimize a loss function determined based on the regularization term and based on a difference between a ground truth image and an image that is estimated according to the volume rendering.