How Do Neural Networks Estimate Optical Flow? A Neuropsychology-Inspired Study

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 44; no. 11; pp. 8290 - 8305
• Main Authors: de Jong, David B., Paredes-Valles, Federico, de Croon, Guido C. H. E.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Artificial neural networks; Biomedical optical imaging; Coders; convolutional neural networks; Encoders-Decoders; Estimation; Flow nets; Gabor filters; Identification methods; Neural networks; neuropsychology; Occlusion; Optical communication; Optical computing; Optical fiber networks; Optical flow; Optical flow (image analysis); Optical imaging; Optical sensors; Pedestrians; Visualization
• ISSN: 0162-8828; 1939-3539; 2160-9292
• DOI: 10.1109/TPAMI.2021.3083538

End-to-end trained convolutional neural networks have led to a breakthrough in optical flow estimation. The most recent advances focus on improving the optical flow estimation by improving the architecture and setting a new benchmark on the publicly available MPI-Sintel dataset. Instead, in this article, we investigate how deep neural networks estimate optical flow. A better understanding of how these networks function is important for (i) assessing their generalization capabilities to unseen inputs, and (ii) suggesting changes to improve their performance. For our investigation, we focus on FlowNetS, as it is the prototype of an encoder-decoder neural network for optical flow estimation. Furthermore, we use a filter identification method that has played a major role in uncovering the motion filters present in animal brains in neuropsychological research. The method shows that the filters in the deepest layer of FlowNetS are sensitive to a variety of motion patterns. Not only do we find translation filters, as demonstrated in animal brains, but thanks to the easier measurements in artificial neural networks, we even unveil dilation, rotation, and occlusion filters. Furthermore, we find similarities in the refinement part of the network and the perceptual filling-in process which occurs in the mammal primary visual cortex.