Normal Similarity Network for Generative Modelling

TL;DR

The paper introduces Normal Similarity Network (NSN), a deep generative model using Gaussian-style filters trained with layer-wise density estimation, capable of generating and reconstructing images effectively.

Contribution

It presents a novel deep generative model with Gaussian-style layers and a non-parametric density training method, along with a new sample generation technique called NSN-Gen.

Findings

Effective image generation, styling, and reconstruction demonstrated

Layer-wise density estimation improves training stability

Model shows promising results across various vision tasks

Abstract

Gaussian distributions are commonly used as a key building block in many generative models. However, their applicability has not been well explored in deep networks. In this paper, we propose a novel deep generative model named as Normal Similarity Network (NSN) where the layers are constructed with Gaussian-style filters. NSN is trained with a layer-wise non-parametric density estimation algorithm that iteratively down-samples the training images and captures the density of the down-sampled training images in the final layer. Additionally, we propose NSN-Gen for generating new samples from noise vectors by iteratively reconstructing feature maps in the hidden layers of NSN. Our experiments suggest encouraging results of the proposed model for a wide range of computer vision applications including image generation, styling and reconstruction from occluded images.