A Study and Comparison of Human and Deep Learning Recognition Performance under Visual Distortions

• Published in: 2017 26th International Conference on Computer Communication and Networks (ICCCN) pp. 1 - 7
• Main Authors: Dodge, Samuel, Karam, Lina
• Format: Conference Proceeding
• Language: English; Japanese
• Published: IEEE 01.07.2017
• Subjects: Distortion; Neural networks; Robustness; Testing; Training; Visual systems
• DOI: 10.1109/ICCCN.2017.8038465

Deep neural networks (DNNs) achieve excellent performance on standard classification tasks. However, under image quality distortions such as blur and noise, classification accuracy becomes poor. In this work, we compare the performance of DNNs with human subjects on distorted images. We show that, although DNNs perform better than or on par with humans on good quality images, DNN performance is still much lower than human performance on distorted images. We additionally find that there is little correlation in errors between DNNs and human subjects. This could be an indication that the internal representation of images are different between DNNs and the human visual system. These comparisons with human performance could be used to guide future development of more robust DNNs.