Implicit Neural Representations

Last updated Nov 2, 2022 Edit Source

Implicit Neural Representations, sometimes also referred to as coordinate-based representations, are a novel way to parameterize signals (of all kinds) as a continuous function that maps the domain of the signal (i.e., a coordinate, such as a pixel coordinate for an image) to whatever is at that coordinate (for an image, an R,G,B color). Of course, these functions are usually not analytically tractable - it is impossible to “write down” the function that parameterizes a natural image as a mathematical formula. Implicit Neural Representations thus approximate that function via a neural network

• https://github.com/vsitzmann/awesome-implicit-representations
• An Introduction to Neural Implicit Representations with Use Cases

• #COURSE Coordinate-based Networks: Implicit Neural Representations

• #TALK TUM AI Lecture Series - Implicit Neural Scene Representations (Vincent Sitzmann)

• #PAPER Implicit Neural Representations with Periodic Activation Functions (Sitzmann 2020)
  • https://www.vincentsitzmann.com/siren/
  • Paper explained
• #PAPER See LIIF in AI/Computer Vision/Super-resolution
• #PAPER Adversarial Generation of Continuous Images (Skorokhodov 2021)
  • https://universome.github.io/inr-gan
• #PAPER Super-Resolution With Local Implicit Image Function and SIREN (Jiang 2021)
• #PAPER Implicit Neural Representation Learning for Hyperspectral Image Super-Resolution (Zhang 2021)
• #PAPER Generalised Implicit Neural Representations (Grattarola 2022)
  • #CODE https://github.com/danielegrattarola/GINR
• #PAPER Towards Generalising Neural Implicit Representations (Costain 2022)
• #PAPER Implicit neural representation for physics-driven actuated soft bodies (Yang 2022)
  • https://www.youtube.com/watch?v=9EERe_CTazk
  • https://studios.disneyresearch.com/2022/07/24/implicit-neural-representation-for-physics-driven-actuated-soft-bodies/