Papers › FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks

6 Dec 2016CVPR 2017 7arXiv:1612.01925archive 2025-07-28

Eddy Ilg, Nikolaus Mayer, Tonmoy Saikia, Margret Keuper, Alexey Dosovitskiy, Thomas Brox

The FlowNet demonstrated that optical flow estimation can be cast as a learning problem. However, the state of the art with regard to the quality of the flow has still been defined by traditional methods. Particularly on small displacements and real-world data, FlowNet cannot compete with variational methods. In this paper, we advance the concept of end-to-end learning of optical flow and make it work really well. The large improvements in quality and speed are caused by three major contributions: first, we focus on the training data and show that the schedule of presenting data during training is very important. Second, we develop a stacked architecture that includes warping of the second image with intermediate optical flow. Third, we elaborate on small displacements by introducing a sub-network specializing on small motions. FlowNet 2.0 is only marginally slower than the original FlowNet but decreases the estimation error by more than 50%. It performs on par with state-of-the-art methods, while running at interactive frame rates. Moreover, we present faster variants that allow optical flow computation at up to 140fps with accuracy matching the original FlowNet.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="1612.01925")

Code

Syntology Ran 2 of 21 code samples harvested from 2 repositories linked to this paper; 19 have no recorded run. Of those that ran: 1 ran · honoured contract; 1 ran · fixture could not drive it.

By repository: community (archive-listed): 21 samples from 2 repositories, 2 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

ElliotHYLee/VisualOdometry3D mentioned on GitHubpytorchMIT report
NVIDIA/flownet2-pytorch mentioned on GitHubpytorch report
lab-midas/lapnet mentioned on GitHubtfMIT report
mcgridles/LENS mentioned on GitHubpytorch report
philferriere/tfoptflow mentioned on GitHubtf report
rickyHong/tfoptflow-repl mentioned on GitHubtf report
simonmeister/UnFlow mentioned on GitHubtf report
open-mmlab/mmflow pytorchApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

21 samples harvested; 2 ran; 1 honoured the contract we drafted; 19 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · honoured contract
1ran · fixture could not drive it
19unverified

Licence: 3 of the 21 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 2 repositories linked to this paper, official or community; each sample names its own and says which. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

compute_color NVIDIA/flownet2-pytorch/utils/flow_utils.py community (archive-listed) ran · honoured contract fingerprinted licence not identified · pointer only · 50f23bdd9e8bbda6 · report
flow2img NVIDIA/flownet2-pytorch/utils/flow_utils.py community (archive-listed) ran · fixture could not drive it fingerprinted licence not identified · pointer only · 025af26f623570d9 · report
arr2kspace lab-midas/lapnet/TF2/core/cropping.py community (archive-listed) unverified MIT (permissive) · 079af00271e054e6 · report
atan2 lab-midas/lapnet/TF2/core/flow_util.py community (archive-listed) unverified MIT (permissive) · f56f923923ad6051 · report
buildLAPNet_model_2D lab-midas/lapnet/TF2/core/lapnet.py community (archive-listed) unverified MIT (permissive) · ac00efb4e7c38ba6 · report
buildLAPNet_model_2D_old lab-midas/lapnet/TF2/core/lapnet.py community (archive-listed) unverified MIT (permissive) · 23bd9ee416efaf7e · report
crop2D_FixPts lab-midas/lapnet/TF2/core/cropping.py community (archive-listed) unverified MIT (permissive) · ac619791f7c058f5 · report
flow_compute_color lab-midas/lapnet/TF2/core/flow_util.py community (archive-listed) unverified MIT (permissive) · 5c88519a582abf72 · report
flow_to_color_np lab-midas/lapnet/TF2/core/flow_util.py community (archive-listed) unverified MIT (permissive) · 201e08769ae9da46 · report
load_mat_file lab-midas/lapnet/TF2/core/util.py community (archive-listed) unverified MIT (permissive) · 54f6f37bc178e63e · report
np_warp_2D lab-midas/lapnet/TF2/core/image_warp.py community (archive-listed) unverified MIT (permissive) · 9ab8dacfcdf8cd27 · report
np_warp_3D lab-midas/lapnet/TF2/core/image_warp.py community (archive-listed) unverified MIT (permissive) · 7dbe406f37174ca7 · report
plot_data lab-midas/lapnet/TF2/core/line_plot.py community (archive-listed) unverified MIT (permissive) · 2c97dc5b158f72ec · report
preprocessing_data lab-midas/lapnet/TF2/core/line_plot.py community (archive-listed) unverified MIT (permissive) · 2b6479fc0d97efd8 · report
readFlow NVIDIA/flownet2-pytorch/utils/flow_utils.py community (archive-listed) unverified licence not identified · pointer only · 1662ea0b62b0229c · report
rectangulartapering2d lab-midas/lapnet/TF2/core/tapering.py community (archive-listed) unverified MIT (permissive) · 2a1f923c7c68c58d · report
rectangulartapering3d lab-midas/lapnet/TF2/core/tapering.py community (archive-listed) unverified MIT (permissive) · dc3bf957592fc5f2 · report
squeeze_func lab-midas/lapnet/TF2/core/lapnet.py community (archive-listed) unverified MIT (permissive) · e0c6c608a0c6f8e3 · report
taper2D lab-midas/lapnet/TF2/core/tapering.py community (archive-listed) unverified MIT (permissive) · 6cc5e9fb591b65e7 · report
test_name_map lab-midas/lapnet/TF1/src/eval_lapnet.py community (archive-listed) unverified MIT (permissive) · def2d3257d021156 · report
warp_assessment3D lab-midas/lapnet/TF2/core/Warp_assessment3D.py community (archive-listed) unverified MIT (permissive) · ca597a74753d4ebe · report

Tasks

Dense Pixel Correspondence EstimationOptical Flow EstimationSkeleton Based Action Recognition

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Dense Pixel Correspondence Estimation HPatches FlowNet2 Viewpoint I AEPE 5.99 #5 of 8 Archive leaderboard report
Dense Pixel Correspondence Estimation HPatches FlowNet2 Viewpoint II AEPE 15.55 #5 of 8 Archive leaderboard report
Dense Pixel Correspondence Estimation HPatches FlowNet2 Viewpoint III AEPE 17.09 #5 of 8 Archive leaderboard report
Dense Pixel Correspondence Estimation HPatches FlowNet2 Viewpoint IV AEPE 22.13 #5 of 8 Archive leaderboard report
Dense Pixel Correspondence Estimation HPatches FlowNet2 Viewpoint V AEPE 30.68 #5 of 8 Archive leaderboard report
Optical Flow Estimation KITTI 2015 (train) FlowNet2 EPE 10.08 #17 of 19 Archive leaderboard report
Optical Flow Estimation KITTI 2015 (train) FlowNet2 F1-all 30.0 #17 of 19 Archive leaderboard report
Optical Flow Estimation Sintel-clean FlowNet2 Average End-Point Error 3.96 #25 of 29 Archive leaderboard report
Optical Flow Estimation Spring FlowNet2 1px total 6.710 #6 of 11 Archive leaderboard report
Skeleton Based Action Recognition JHMDB Pose Tracking FlowNet2 PCK@0.1 45.2 #3 of 3 Archive leaderboard report
Skeleton Based Action Recognition JHMDB Pose Tracking FlowNet2 PCK@0.2 62.9 #3 of 3 Archive leaderboard report
Skeleton Based Action Recognition JHMDB Pose Tracking FlowNet2 PCK@0.3 73.5 #3 of 3 Archive leaderboard report
Skeleton Based Action Recognition JHMDB Pose Tracking FlowNet2 PCK@0.4 80.6 #3 of 3 Archive leaderboard report
Skeleton Based Action Recognition JHMDB Pose Tracking FlowNet2 PCK@0.5 85.5 #3 of 3 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

SPEED

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections