Weakly-Supervised Optical Flow Estimation for Time-of-Flight

TL;DR

This paper introduces a novel training method for optical flow networks that directly supervises on reconstructed depth data from iToF sensors, effectively reducing motion artifacts without requiring ground truth flow labels.

Contribution

It presents a new training algorithm for optical flow that leverages depth supervision from iToF data, enabling motion artifact correction without ground truth flow annotations.

Findings

Outperforms existing motion compensation methods.

Effective for single- and multi-frequency iToF sensors.

Reduces motion artifacts in depth images.

Abstract

Indirect Time-of-Flight (iToF) cameras are a widespread type of 3D sensor, which perform multiple captures to obtain depth values of the captured scene. While recent approaches to correct iToF depths achieve high performance when removing multi-path-interference and sensor noise, little research has been done to tackle motion artifacts. In this work we propose a training algorithm, which allows to supervise Optical Flow (OF) networks directly on the reconstructed depth, without the need of having ground truth flows. We demonstrate that this approach enables the training of OF networks to align raw iToF measurements and compensate motion artifacts in the iToF depth images. The approach is evaluated for both single- and multi-frequency sensors as well as multi-tap sensors, and is able to outperform other motion compensation techniques.