We question the dominant role of real-world training images in the field of material classification by investigating whether synthesized data can generalise more effectively than real-world data. Experimental results on three challenging real-world material databases show that the best performing pre-trained convolutional neural network (CNN) architectures can achieve up to 91.03% mean average precision when classifying materials in cross-dataset scenarios. We demonstrate that synthesized data achieve an improvement on mean average precision when used as training data and in conjunction with pre-trained CNN architectures, which spans from ~ 5% to ~ 19% across three widely used material databases of real-world images.
@article{arxiv.1711.03874,
title = {Material Classification in the Wild: Do Synthesized Training Data Generalise Better than Real-World Training Data?},
author = {Grigorios Kalliatakis and Anca Sticlaru and George Stamatiadis and Shoaib Ehsan and Ales Leonardis and Juergen Gall and Klaus D. McDonald-Maier},
journal= {arXiv preprint arXiv:1711.03874},
year = {2017}
}
Comments
accepted for publication in VISAPP 2018. arXiv admin note: text overlap with arXiv:1703.04101