Papers › A Closer Look at Spatiotemporal Convolutions for Action Recognition

A Closer Look at Spatiotemporal Convolutions for Action Recognition

30 Nov 2017CVPR 2018 6arXiv:1711.11248archive 2025-07-28

Du Tran, Heng Wang, Lorenzo Torresani, Jamie Ray, Yann Lecun, Manohar Paluri

In this paper we discuss several forms of spatiotemporal convolutions for video analysis and study their effects on action recognition. Our motivation stems from the observation that 2D CNNs applied to individual frames of the video have remained solid performers in action recognition. In this work we empirically demonstrate the accuracy advantages of 3D CNNs over 2D CNNs within the framework of residual learning. Furthermore, we show that factorizing the 3D convolutional filters into separate spatial and temporal components yields significantly advantages in accuracy. Our empirical study leads to the design of a new spatiotemporal convolutional block "R(2+1)D" which gives rise to CNNs that achieve results comparable or superior to the state-of-the-art on Sports-1M, Kinetics, UCF101 and HMDB51.

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="1711.11248")

Code

Syntology Ran 1 of 4 code samples harvested from 2 repositories linked to this paper; 3 have no recorded run. Of those that ran: 1 ran · our draft was wrong.

By repository: community (archive-listed): 4 samples from 2 repositories, 1 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

24 repositories listed; official and paper-mentioned ones first.

facebookresearch/R2Plus1D officialmentioned in papermentioned on GitHubcaffe2Apache-2.0 report
2023-MindSpore-1/ms-code-187 mentioned on GitHubmindspore report
2023-MindSpore-1/ms-code-68 mentioned on GitHubmindspore report
3dperceptionlab/visual-wetlandbirds mentioned on GitHubpytorchNOASSERTION report
AD2605/Action-Recognition mentioned on GitHubpytorch report
Bangbangbanana/r2plus1d_mindspore mentioned on GitHubmindspore report
BelixRogner/SpeedChallenge mentioned on GitHubpytorch report
JuliBaCSE/WeCare_makeathon mentioned on GitHubpytorch report
facebookresearch/VMZ mentioned on GitHubcaffe2 report
fmthoker/severe-benchmark mentioned on GitHubpytorch report
juenkhaw/action_recognition_project mentioned on GitHubpytorch report
juenkhaw/action_recognition_project1 mentioned on GitHubpytorch report
karatuno/Action-Recognition mentioned on GitHubtf report
kietngt00/hmdb51-recognition mentioned on GitHubpytorch report
leftthomas/r2plus1d-c3d mentioned on GitHubpytorch report
wasilone11/ICPR-RIP-2024 mentioned on GitHub 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

4 samples harvested; 1 ran; 0 honoured the contract we drafted; 3 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 · our draft was wrong
3unverified

Licence: 4 of the 4 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.

r3d_18 kietngt00/hmdb51-recognition/model/R2plus1D.py community (archive-listed) ran · our draft was wrong no licence file found · pointer only · 3c34bc53cbb7d79d · report
center_crop leftthomas/r2plus1d-c3d/inference.py community (archive-listed) unverified no licence file found · pointer only · b09d9df7736daaa3 · report
mc3_18 kietngt00/hmdb51-recognition/model/R2plus1D.py community (archive-listed) unverified no licence file found · pointer only · b0848b1c8d35150b · report
r2plus1d_18 kietngt00/hmdb51-recognition/model/R2plus1D.py community (archive-listed) unverified no licence file found · pointer only · 8fb2663dede3c507 · report

Tasks

Action ClassificationAction RecognitionTemporal Action Localization

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Action Classification Kinetics-400 R[2+1]D-Flow (Sports-1M pretrain) Acc@1 75.4 #156 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-Flow (Sports-1M pretrain) Acc@5 91.9 #156 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-RGB (Sports-1M pretrain) Acc@1 74.3 #163 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-RGB (Sports-1M pretrain) Acc@5 91.4 #163 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-Two-Stream Acc@1 73.9 #165 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-Two-Stream Acc@5 90.9 #165 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D Acc@1 72 #175 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D Acc@5 90 #175 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-RGB Acc@1 72 #176 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-RGB Acc@5 90 #176 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-Flow Acc@1 67.5 #185 of 207 Archive leaderboard report
Action Classification Kinetics-400 R[2+1]D-Flow Acc@5 87.2 #185 of 207 Archive leaderboard report
Action Recognition HMDB-51 R[2+1]D-TwoStream (Kinetics pretrained) Average accuracy of 3 splits 78.7 #28 of 77 Archive leaderboard report
Action Recognition HMDB-51 R[2+1]D-Flow (Kinetics pretrained) Average accuracy of 3 splits 76.4 #36 of 77 Archive leaderboard report
Action Recognition HMDB-51 R[2+1]D-RGB (Kinetics pretrained) Average accuracy of 3 splits 74.5 #41 of 77 Archive leaderboard report
Action Recognition HMDB-51 R[2+1D]D-TwoStream (Sports1M pretrained) Average accuracy of 3 splits 72.7 #46 of 77 Archive leaderboard report
Action Recognition HMDB-51 R[2+1]D-Flow (Sports1M pretrained) Average accuracy of 3 splits 70.1 #55 of 77 Archive leaderboard report
Action Recognition HMDB-51 R[2+1]D-RGB (Sports1M pretrained) Average accuracy of 3 splits 66.6 #59 of 77 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-Two-Stream-32frame Video hit@1 73.3 #3 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-Two-Stream-32frame Video hit@5 91.9 #3 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-RGB-32frame Clip Hit@1 57 #4 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-RGB-32frame Video hit@1 73 #4 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-RGB-32frame Video hit@5 91.5 #4 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-Flow-32frame Clip Hit@1 46.4 #6 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-Flow-32frame Video hit@1 68.4 #6 of 9 Archive leaderboard report
Action Recognition Sports-1M R[2+1]D-Flow-32frame Video hit@5 88.7 #6 of 9 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-TwoStream (Kinetics pretrained) 3-fold Accuracy 97.3 #20 of 91 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-RGB (Kinetics pretrained) 3-fold Accuracy 96.8 #30 of 91 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-Flow (Kinetics pretrained) 3-fold Accuracy 95.5 #44 of 91 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-TwoStream (Sports-1M pretrained) 3-fold Accuracy 95 #49 of 91 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-RGB (Sports-1M pretrained) 3-fold Accuracy 93.6 #59 of 91 Archive leaderboard report
Action Recognition UCF101 R[2+1]D-Flow (Sports-1M pretrained) 3-fold Accuracy 93.3 #61 of 91 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

Introduced by this paper: (2+1)D Convolution, R(2+1)D

(2+1)D ConvolutionAverage PoolingBatch NormalizationDense ConnectionsGlobal Average PoolingLinear WarmupR(2+1)DRandom Resized CropReLUResidual ConnectionSGD with MomentumTemporal JitteringWeight Decay

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