README.md

PWC-Net

PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume

Abstract

We present a compact but effective CNN model for optical flow, called PWC-Net. PWC-Net has been designed according to simple and well-established principles: pyramidal processing, warping, and the use of a cost volume. Cast in a learnable feature pyramid, PWC-Net uses the current optical flow estimate to warp the CNN features of the second image. It then uses the warped features and features of the first image to construct a cost volume, which is processed by a CNN to estimate the optical flow. PWC-Net is 17 times smaller in size and easier to train than the recent FlowNet2 model. Moreover, it outperforms all published optical flow methods on the MPI Sintel final pass and KITTI 2015 benchmarks, running at about 35 fps on Sintel resolution (1024×436) images. Our models are available on https://github.com/NVlabs/PWC-Net.

Results and Models

Models	Training datasets	FlyingChairs	Sintel (training)		KITTI2012 (training)	KITTI2015 (training)		Log	Config	Download
			clean	final	EPE	Fl-all	EPE
PWC-Net	FlyingChairs	1.51	3.52	4.81	-	-	-	log	config	model
PWC-Net	Flying Chairs + FlyingThing3d subset	-	2.26	3.79	3.66	29.85%	9.49	log	config	model
PWC-Net-ft	Flying Chairs + FlyingThing3d subset + Sintel	-	1.50	2.06	-	-	-	log	config	model
PWC-Net-ft-final	FlyingChairs + FlyingThing3d subset + Sintel final	-	1.82	1.78	-	-	-	log	config	model
PWC-Net-ft	FlyingChairs + FlyingThing3d subset + KITTI	-	-	-	1.07	6.09%	1.64	log	config	model
PWC-Net+	FlyingChairs + FlyingThing3d subset + Sintel + KITTI2015 + HD1K	-	1.90	2.39	-	-	-	log	config	model

Citation

@inproceedings{sun2018pwc,
  title={Pwc-net: Cnns for optical flow using pyramid, warping, and cost volume},
  author={Sun, Deqing and Yang, Xiaodong and Liu, Ming-Yu and Kautz, Jan},
  booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
  pages={8934--8943},
  year={2018}
}