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

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...

How do deep neural networks perform optical flow estimation?: A neuropsychology-inspired study

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...

Vector Doppler imaging of small vessels using directionally filtered Power Doppler images

Power Doppler (PD) imaging has become a staple in high frame rate ultrasound imaging due to its ability to image small vessels and slow-moving flows, such as in the case of imaging blood flow in the brain. Alternatively, color Doppler (CD) can be used to determine the one-dimensional directional information of the blood scatterers. This can help...