Coordinate-based Multilayer Perceptron (MLP) networks, despite being capable of learning neural implicit representations, are not performant for internal image synthesis applications. Convolutional Neural Networks (CNNs) are typically used instead for a variety of internal generative tasks, at the cost of a larger model. We propose Neural Knitwork, an architecture for neural implicit representation learning of natural images that achieves image synthesis by optimizing the distribution of image patches in an adversarial manner and by enforcing consistency between the patch predictions. To the best of our knowledge, this is the first implementation of a coordinate-based MLP tailored for synthesis tasks such as image inpainting, super-resolution, and denoising. We demonstrate the utility of the proposed technique by training on these three tasks. The results show that modeling natural images using patches, rather than pixels, produces results of higher fidelity. The resulting model requires 80% fewer parameters than alternative CNN-based solutions while achieving comparable performance and training time.
Original languageEnglish
Place of PublicationIthaca, N.Y.
Number of pages11
Publication statusPublished - 29 Sep 2021


  • multilayer perceptron (MLP)
  • convolutional neural network (CNN)
  • neural knitwork


Dive into the research topics of 'Neural knitworks: patched neural implicit representation networks'. Together they form a unique fingerprint.

Cite this