presented at ACM Siggraph 2020 (Posters)

The role of objective and subjective measures in material similarity learning

Johanna Delanoy1 Manuel Lagunas1 Ignacio Galve1 Diego Gutierrez1 Ana Serrano1 Roland Fleming2 Belen Masia1
1Universidad de Zaragoza, I3A 2University of Giessen


Establishing a robust measure for material similarity that correlates well with human perception is a long-standing problem. A recent work presented a deep learning model trained to produce a feature space that aligns with human perception by gathering human subjective measures. The resulting metric outperforms objective existing ones. In this work, we aim to understand whether this increased performance is a result of using human perceptual data or is due to the nature of feature learnt by deep learning models. We train similar networks with objective measures (BRDF similarity or classification task) and show that these networks can predict human judgements as well, suggesting that the non-linear features learnt by convolutional network might be a key to model material perception.


	author = {Delanoy, Johanna and Lagunas, Manuel and Galve, Ignacio and Gutierrez, Diego and Serrano, Ana and Fleming, Roland and Masia, Belen},
	title = {The role of objective and subjective measures in material similarity learning},
	booktitle = {ACM SIGGRAPH 2020 Posters},
	year = {2020},
	publisher = {Association for Computing Machinery},
	url = {},
	doi = {},