OmniFlow: Human Omnidirectional Optical Flow
This provides a new dataset for researchers in computer vision to improve optical flow estimation, but it is incremental as it focuses on synthetic data generation rather than novel algorithmic advances.
The paper introduces OmniFlow, a synthetic omnidirectional human optical flow dataset with 23,653 image pairs, created using a rendering engine to simulate naturalistic 3D indoor environments for training optical flow estimation networks.
Optical flow is the motion of a pixel between at least two consecutive video frames and can be estimated through an end-to-end trainable convolutional neural network. To this end, large training datasets are required to improve the accuracy of optical flow estimation. Our paper presents OmniFlow: a new synthetic omnidirectional human optical flow dataset. Based on a rendering engine we create a naturalistic 3D indoor environment with textured rooms, characters, actions, objects, illumination and motion blur where all components of the environment are shuffled during the data capturing process. The simulation has as output rendered images of household activities and the corresponding forward and backward optical flow. To verify the data for training volumetric correspondence networks for optical flow estimation we train different subsets of the data and test on OmniFlow with and without Test-Time-Augmentation. As a result we have generated 23,653 image pairs and corresponding forward and backward optical flow. Our dataset can be downloaded from: https://mytuc.org/byfs