• Corpus ID: 31022778

TransFlow: Unsupervised Motion Flow by Joint Geometric and Pixel-level Estimation

  title={TransFlow: Unsupervised Motion Flow by Joint Geometric and Pixel-level Estimation},
  author={Stefano Alletto and Davide Abati and Simone Calderara and Rita Cucchiara and Luca Rigazio},
We address unsupervised optical flow estimation for ego-centric motion. We argue that optical flow can be cast as a geometrical warping between two successive video frames and devise a deep architecture to estimate such transformation in two stages. First, a dense pixel-level flow is computed with a geometric prior imposing strong spatial constraints. Such prior is typical of driving scenes, where the point of view is coherent with the vehicle motion. We show how such global transformation can… 

Figures and Tables from this paper

Self-Superflow: Self-Supervised Scene Flow Prediction in Stereo Sequences

This work explores the extension of a self-supervised loss based on the Census transform and occlusion-aware bidirectional displacements for the problem of scene flow prediction and shows improved generalization capabilities while achieving much faster convergence.

Unsupervised Learning of Multi-Frame Optical Flow with Occlusions

This paper exploits the minimal configuration of three frames to strengthen the photometric loss and explicitly reason about occlusions and demonstrates that their multi-frame, occlusion-sensitive formulation outperforms existing unsupervised two-frame methods and even produces results on par with some fully supervised methods.

Temporal Interpolation as an Unsupervised Pretraining Task for Optical Flow Estimation

This work investigates frame interpolation as a proxy task for optical flow using real movies and trains a CNN unsupervised for temporal interpolation, which outperforms similar architectures that were trained supervised using synthetic optical flow.

Optical Flow Estimation with Deep Learning, a Survey on Recent Advances

This chapter begins with reviewing traditional (handcrafted) approaches, then introduces the more recent approaches, and finally gets concluded with surveying deep learning approaches.

Self-Supervised Optical Flow Learning

The experiments regarding semisupervision show that including the unsupervised objective with the proposed method significantly improves the estimation on a distant domain while maintaining the performance on the original domain, and the results suggest that the methods are unable to accurately detect occlusions.

Comparative Study of latest CNN based Optical Flow Estimation

This paper has tried off extensive analyses and categorized various deep learning-based optical flow estimation techniques and identified the differences and the correspondences between deep methods and conventional methods.

Proceedings of the 25th Computer Vision Winter Workshop Conference

Vehicle detection in aerial and satellite images is still challenging due to their tiny appearance in pixels compared to the overall size of remote sensing imagery. Classical methods of object

Exploiting Semantic Information and Deep Matching for Optical Flow

This work tackles the problem of estimating optical flow from a monocular camera in the context of autonomous driving by proposing to estimate the traffic participants using instance-level segmentation, and introduces a new convolutional net that learns to perform flow matching, and is able to estimates the uncertainty of its matches.

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

The concept of end-to-end learning of optical flow is advanced and it work really well, and faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet are presented.

Large Displacement Optical Flow: Descriptor Matching in Variational Motion Estimation

  • T. BroxJitendra Malik
  • Computer Science
    IEEE Transactions on Pattern Analysis and Machine Intelligence
  • 2011
A way to approach the problem of dense optical flow estimation by integrating rich descriptors into the variational optical flow setting, while reaching out to new domains of motion analysis where the requirement of dense sampling in time is no longer satisfied is presented.

Back to Basics: Unsupervised Learning of Optical Flow via Brightness Constancy and Motion Smoothness

An unsupervised approach to train a convnet end-to-end for predicting optical flow between two images using a loss function that combines a data term that measures photometric constancy over time with a spatial term that models the expected variation of flow across the image.

EpicFlow: Edge-preserving interpolation of correspondences for optical flow

The proposed approach, called Edge-Preserving Interpolation of Correspondences (EpicFlow) is fast and robust to large displacements, and significantly outperforms the state of the art on MPI-Sintel and performs on par on Kitti and Middlebury.

High Accuracy Optical Flow Estimation Based on a Theory for Warping

By proving that this scheme implements a coarse-to-fine warping strategy, this work gives a theoretical foundation for warping which has been used on a mainly experimental basis so far and demonstrates its excellent robustness under noise.

DeepFlow: Large Displacement Optical Flow with Deep Matching

This work proposes a descriptor matching algorithm, tailored to the optical flow problem, that allows to boost performance on fast motions, and sets a new state-of-the-art on the MPI-Sintel dataset.

Large Displacement Optical Flow from Nearest Neighbor Fields

An optical flow algorithm for large displacement motions that uses approximate nearest neighbor fields to compute an initial motion field and a robust algorithm to compute a set of similarity transformations as the motion candidates for segmentation to account for deviations from similarity transformations.

VirtualWorlds as Proxy for Multi-object Tracking Analysis

This work proposes an efficient real-to-virtual world cloning method, and validate the approach by building and publicly releasing a new video dataset, called "Virtual KITTI", automatically labeled with accurate ground truth for object detection, tracking, scene and instance segmentation, depth, and optical flow.

Determining Optical Flow

An iterative implementation is shown which successfully computes the optical flow for a number of synthetic image sequences and is robust in that it can handle image sequences that are quantified rather coarsely in space and time.