4 datasets found

The FSVG dataset is created by collecting and labeling an enormous amount of data from videogames such as GTAV and The Witcher 3. It includes densely labeled depth, optical...

The dataset used for training and testing the proposed deep single-image intrinsic decomposition model.

Albedo, shading, and shading detail datasets used to train the conv-VAE models.

A seminal work on Neural Painting, proposing heuristic-based methods to decompose a given image into a set of strokes.