CVFeb 22, 2017
Synthesising Dynamic Textures using Convolutional Neural Networks
arXiv:1702.07006v118 citations
AI Analysis
This work addresses the problem of dynamic texture synthesis for computer vision applications, but it appears incremental as it builds on existing CNN methods.
The authors tackled the problem of modeling dynamic textures by introducing a parametric model based on spatiotemporal summary statistics from CNN features, enabling synthesis of new texture samples and motion prediction in simple movies.
Here we present a parametric model for dynamic textures. The model is based on spatiotemporal summary statistics computed from the feature representations of a Convolutional Neural Network (CNN) trained on object recognition. We demonstrate how the model can be used to synthesise new samples of dynamic textures and to predict motion in simple movies.