Parametric Encoding
Parametric encoding represents data using a set of parameters, aiming to achieve efficient and effective data representation for various tasks. Current research focuses on developing novel parametric encoding schemes tailored to specific data types, such as meshes, audio signals, and images, often integrating these with neural networks (e.g., convolutional, attention-based architectures) to improve performance. This approach is proving valuable across diverse fields, enhancing the speed and quality of applications ranging from real-time 3D scene reconstruction and audio compression to solving partial differential equations and quantum-enhanced image processing.
Papers
July 18, 2024
May 23, 2024
March 22, 2024
February 16, 2024
July 25, 2023
April 27, 2023
April 7, 2023
November 8, 2022