The DOI matches its Crossref record; 27 of the 27 references it cites were found in a registry or an index.
A Duality Based Approach for Realtime TV-L 1 Optical Flow
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“A reasonable explanation is that the temporal network is actually designed to capture different visual aspects of videos compared to the spatial network, which indicates that the feature distributions are more diverse between those fields (e.g., optical flow and RGB) [20], [24], [25].”
From Unsupervised Deep Video Hashing via Balanced Code for Large-Scale Video Retrieval (2019) · cites this paper as [24]
Citations
Cites
Nonlinear total variation based noise removal algorithms
Rudin, L., Osher, S., & Fatemi, E. (1992) · Physica D Nonlinear Phenomena · doi:10.1016/0167-2789(92)90242-f · Cited by 17,214
Horn, B. K. P., & Schunck, B. G. (1981) · Artificial Intelligence · doi:10.1016/0004-3702(81)90024-2 · Cited by 10,312
Horn, B. K. P., & Schunck, B. G. (1981) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.965761 · Cited by 7,443
chambolle (2004) · doi:10.1023/b:jmiv.0000011321.19549.88 · Cited by 3,227
High Accuracy Optical Flow Estimation Based on a Theory for Warping
Brox, T., Bruhn, A., Papenberg, N., & Weickert, J. (2004) · Lecture notes in computer science · doi:10.1007/978-3-540-24673-2_3 · Cited by 2,634
A computational framework and an algorithm for the measurement of visual motion
Anandan, P. (1989) · International Journal of Computer Vision · doi:10.1007/bf00158167 · Cited by 1,073
Thormaehlen, T., Broszio, H., & Weißenfeld, A. (2004) · Lecture notes in computer science · doi:10.1007/b97865 · Cited by 901
Fast Global Minimization of the Active Contour/Snake Model
Bresson, X., Esedoḡlu, S., Vandergheynst, P., Thiran, J., & Osher, S. (2007) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-007-0002-0 · Cited by 857
A Nonlinear Primal-Dual Method for Total Variation-Based Image Restoration
Chan, T. F.-C., Golub, G. H., & Mulet, P. (1999) · SIAM Journal on Scientific Computing · doi:10.1137/s1064827596299767 · Cited by 849
An Algorithm for Total Variation Minimization and Applications
Chambolle, A. (2004) · Journal of Mathematical Imaging and Vision · doi:10.1023/b:jmiv.0000011325.36760.1e · Cited by 835
Cited by
Sorted by citations
Items 901–1,000 of 1,002
On-Board Monocular Vision System Pose Estimation through a Dense Optical Flow
Onkarappa, N., & Sappa, A. D. (2010) · Lecture notes in computer science · doi:10.1007/978-3-642-13772-3_24 · Cited by 7
“Since the terms in (1) are not continuously dif f erentiable, the energy function can be minimized using dual formulation for minimizing total variation as proposed in [16] and adapted to optical flow in [ 13 ].”
Method3D Urban Reconstruction from Wide Area Aerial Surveillance Video
Kang, Z., & Medioni, G. G. (2015) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw.2015.17 · Cited by 7
“The best and most recent method to compute optical flow is the Total-Variation L1 (TV-L1) optical flow as described in [26], which minimizes an L1 norm data term and a totalvariation regularization term:”
MethodSolving Quasi-Variational Inequalities for Image Restoration with Adaptive Constraint Sets
Lenzen, F., Lellmann, J., Becker, F., & Schnörr, C. (2014) · SIAM Journal on Imaging Sciences · doi:10.1137/130938347 · Cited by 7
“, for optical flow [41, 40] and multilabeling [21, 37].”
BackgroundFast L1 smoothing splines with an application to Kinect depth data
Tepper, M., & Sapiro, G. (2013) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2013.6738104 · Cited by 7
A Coding-Cost Framework for Super-Resolution Motion Layer Decomposition
Schoenemann, T., & Cremers, D. (2011) · IEEE Transactions on Image Processing · doi:10.1109/tip.2011.2169271 · Cited by 7
A high-performance parallel implementation of the Chambolle algorithm
Akın, A., Beretta, I., Nacci, A. A., Rana, V., Santambrogio, M. D., & Atienza, D. (2011) · Design, Automation, and Test in Europe · doi:10.1109/date.2011.5763232 · Cited by 7
“Firstly, a support variable v = (v1, v2) is defined using a thresholding function of I1 and of the value of u computed at the previous level [13].”
BackgroundMethodInfluentialExploiting High Level Scene Cues in Stereo Reconstruction
Hadfield, S., & Bowden, R. (2015) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.96 · Cited by 7
“Various approaches to encoding these spatial smoothness have been proposed including: Total Variation (L2 and L1 [39, 24]), Monte-carlo inspired PatchMatch approaches [3, 17] and the Total Generalized Variation [31, 25] which helps overcome the staircasing artifacts caused by total variation regularisation.”
MethodStereo reconstruction using top-down cues
Hadfield, S., Lebeda, K., & Bowden, R. (2016) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2016.08.001 · Cited by 7
“Some examples of these spatial-smoothness constraints include Total Variation (using both L1 and L2 norms [22, 23]), Monte-carlo inspired PatchMatch approaches [24, 25] and the Total Generalized Variation [26, 27] which was designed to deal with staircasing artifacts which are often introduced by total variation techn…”
The differential method for calculation of cloud motion vectors
Simonenko, E. V., Chudin, A. O., & Николаевич, Д. А. (2017) · Russian Meteorology and Hydrology · doi:10.3103/s1068373917030037 · Cited by 7
A dynamic causal topic model for mining activities from complex videos
Fan, Y., Zhou, Q., Yue, W., & Zhu, W. (2017) · Multimedia Tools and Applications · doi:10.1007/s11042-017-4760-4 · Cited by 7
“TV-L1 algorithm [27] is used to extract optical flow vectors from each frame.”
MethodIterative Refinement of Transmission Map for Stereo Image Defogging Using a Dual Camera Sensor
Kim, H., Park, J., Park, H., & Paik, J. (2017) · Sensors · doi:10.3390/s17122861 · Cited by 7
“To minimize the CLG–TV error function, we use an alternative Horn–Schunck model [29] where ECLG−TV is decomposed into three terms, as shown below.”
MethodDeep spectral feature pyramid in the frequency domain for long-term action recognition
An, G., Zheng, Z., Wu, D. O., & Zhou, W. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102650 · Cited by 7
A High-Performance Dense Optical Flow Architecture Based on Red-Black SOR Solver
Johnson, B., Thomas, S., & Rani, J. S. (2019) · Journal of Signal Processing Systems · doi:10.1007/s11265-019-01490-5 · Cited by 7
“[4] proposed a modified HSOF algorithm on NVIDIA GeForce 7800 GS and GeForce Go 7900 GTX GPGPU to compute dense OF from Yosemite image (316 × 252) sequence at about Journal of Signal Processing Systems (2020) 92:357–373”
BackgroundMethodAn Improved Attention-Based Spatiotemporal-Stream Model for Action Recognition in Videos
Liu, D., Ji, Y., Ye, M., Gan, Y., & Zhang, J. (2020) · IEEE Access · doi:10.1109/access.2020.2983355 · Cited by 7
“In order to learn the temporal information well, we first get the optical flow videos from original action videos [37].”
BackgroundReal-time Detection of Activities in Untrimmed Videos
Gleason, J., Castillo, C. D., & Chellappa, R. (2020) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw50321.2020.9096937 · Cited by 7
“Optical flow is computed using the TV-L1 algorithm [28].”
MethodMutual modality learning for video action classification
Komkov, S. A., Dzabraev, M., & Petiushko, A. (2023) · Computer Optics · doi:10.18287/2412-6179-co-1277 · Cited by 8
Condensing a Sequence to One Informative Frame for Video Recognition
Qiu, Z., Yao, T., Shu, Y., Ngo, C., & Mei, T. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01600 · Cited by 7
“The optical flow modality firstly computes the value of optical flow between consecutive frames with [46], and then converts to a 2-channel image by scaling the value to [0, 255].”
Implementation of VBM3D and Some Variants
Ehret, T., & Arias, P. (2021) · Image Processing On Line · doi:10.5201/ipol.2021.340 · Cited by 7
Audio-Visual Temporal Saliency Modeling Validated by fMRI Data
Koutras, P., Panagiotaropoulou, G., Tsiami, A., & Maragos, P. A. (2018) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2018.00269 · Cited by 7
“For temporal saliency, we extract warped optical flow maps using the implementation of [67], which is based on the TVL1 optical flow algorithm [70].”
MethodAdaptive Regularization of Some Inverse Problems in Image Analysis
Hong, B., Koo, J., Burger, M., & Soatto, S. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2960587 · Cited by 7
Temporal Feature Weighting for Prototype-Based Action Recognition
Mauthner, T., Roth, P. M., & Bischof, H. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-19309-5_44 · Cited by 7
“In particular, to estimate the optical dense flow field, we apply the method proposed in [15], which is publicly available via OFLib.”
MethodSelf-citationCombining Contrast Invariant L1 Data Fidelities with Nonlinear Spectral Image Decomposition
Zeune, L. L., van Gils, S. A., Terstappen, L. W. M. M., & Brüne, C. (2017) · Lecture notes in computer science · doi:10.1007/978-3-319-58771-4_7 · Cited by 7
“For this reason, L 1 data fidelities have been successfully used for advanced image reconstruction techniques [12,13,14], vector field estimation [15] and image decompositions regarding texture [10,16].”
MethodWho's Better, Who's Best: Skill Determination in Video using Deep Ranking.Preprint
Doughty, H., Damen, D., & Mayol-Cuevas, W. W. (2017) · arXiv (Cornell University) · Cited by 7
IF-TTN: Information Fused Temporal Transformation Network for Video Action RecognitionPreprint
Yang, K., Qiao, P., Li, D., & Dou, Y. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.09928 · Cited by 7
“For optical flow extraction, we use TVL1 algorithm [31] implemented in OpenCV with CUDA.”
MethodTowards Training Stronger Video Vision Transformers for EPIC-KITCHENS-100 Action RecognitionPreprint
Huang, Z., Qing, Z., Xiang, W., Feng, Y., Zhang, S., Jiang, J., Xia, Z., Tang, M., Sang, N., & Ang, M. H. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.05058 · Cited by 7
“We also trained ViViT on the optical flow of Kinetics 400, which we extract using Raft [15] and TVL1 [21].”
MethodATG-PVD: Ticketing Parking Violations on A DronePreprint
Wang, H., Liu, Y., Huang, H., Pan, Y., Yu, W., Jiang, J., Lyu, D., Bocus, M. J., Liu, M., Pitas, I., et al. (2020) · doi:10.36227/techrxiv.12839156 · Cited by 7
“Traditional approaches generally formulate optical flow estimation as a global energy minimization problem [8,9,10,11].”
BackgroundAnalyzing the Visual Road Scene for Driver Stress Estimation
Bustos, C., Solé‐Ribalta, A., Elhaouij, N., Borge‐Holthoefer, J., Lapedriza, À., & Picard, R. W. (2025) · IEEE Transactions on Affective Computing · doi:10.1109/taffc.2025.3539003 · Cited by 6
“Wewouldalso like to mention that we conducted tests involving optical flow data [67], derived from the sequence of frames, within additional branches of the TSN model.”
FlowSeek: Optical Flow Made Easier with Depth Foundation Models and Motion Bases
Poggi, M., & Tosi, F. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00537 · Cited by 6
“Optical flow estimation has evolved from classical approaches that treated it as an optimization problem [3, 22, 95] to modern deep learning methods.”
Foreground prediction for bilayer segmentation of videos
Tang, Z., Miao, Z., Wan, Y., & Jesse, F. F. (2011) · Pattern Recognition Letters · doi:10.1016/j.patrec.2011.07.025 · Cited by 6
“We use an optical flow algorithm (Zach et al., 2007) to achieve this goal.”
MethodAli, S., Daul, C., & Blondel, W. (2014) · International Conference on Image Processing Theory, Tools and Applications (IPTA) · doi:10.1109/ipta.2014.7001947 · Cited by 6
Hahn, M., Barrois, B., Krüger, L., Wöhler, C., Sagerer, G., & Kümmert, F. (2010) · 3D Research · doi:10.1007/3dres.03(2010)03 · Cited by 6
Preserving accurate motion contours with reliable parameter selection
Sánchez, J., Salgado, A., & Monzón, N. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025041 · Cited by 6
“[14].”
Weakly-supervised Temporal Action Localization with Adaptive Clustering and Refining Network
Ren, H., Ran, W., Liu, X., Ren, H., Lu, H., Zhang, R., & Jin, C. (2023) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme55011.2023.00177 · Cited by 6
“The optical flow is extracted by TV-L1 algorithm [21].”
MethodCorrelation Image Sensor-Assisted Single-Frame Optical Flow Estimation in Motion-Blurred Scenes
Wang, P., Kurihara, T., & Yu, J. (2024) · IEEE Access · doi:10.1109/access.2024.3398399 · Cited by 6
“Despite the widespread use of two-frame methods [9], [10], [11], [12], [13], [14], which use two consecutive frames in a video to estimate optical flow, these methods encounter limitations when dealing with scenes afflicted by motion blur.”
MethodAn Evaluation Approach for Scene Flow with Decoupled Motion and Position
Wedel, A., Vaudrey, T., Meißner, A., Rabe, C., Brox, T., Franke, U., & Cremers, D. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-03061-1_3 · Cited by 6
“Real-time methods have been proposed in [3,24].”
Brkić, K., Rašić, S., Pinz, A. J., Šegvić, S., & Kalafatić, Z. (2014) · Proceedings of the Croatian Computer Vision Workshop · doi:10.20532/ccvw.2013.0007 · Cited by 6
Hui, T.-W., & Chung, R. H. Y. (2013) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2013.04.007 · Cited by 6
Robust bladder image registration by redefining data-term in total variational approach
Ali, S., Daul, C., Galbrun, E., Amouroux, M., Guillemin, F. H., & Blondel, W. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2077658 · Cited by 6
“p = p + σ∇L̃ max(1, |pn + σ∇L̃n|) q = q + σ∇ũ max(1, |qn + σ∇ũ|) dual variable update (22)”
Interactive Scene Flow Editing for Improved Image-based Rendering and Virtual Spacetime Navigation
Ruhl, K., Eisemann, M., Hilsmann, A., Eisert, P., & Magnor, M. A. (2015) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/2733373.2806251 · Cited by 6
“With respect to the latter, our scene flow optimization could potentially benefit from a primaldual approach in the style of [22], left for future work.”
BackgroundImplementation of Local Fourier Burst Accumulation for Video Deblurring
Anger, J., & Meinhardt-Llopis, E. (2017) · Image Processing On Line · doi:10.5201/ipol.2017.197 · Cited by 6
Collateral motion saliency-based model for Trypanosoma cruzi detection in dye-free blood microscopy
Martins, G. L., Ferreira, D. S., & Ramalho, G. L. B. (2021) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2021.104220 · Cited by 6
Poggi, M., Aleotti, F., & Mattoccia, S. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00781 · Cited by 6
“Since the seminal work by Horn and Schunck [21], for years optical flow has been cast into an energy minimization problem [8, 7, 9, 60, 59], for instance by means of variational frameworks [6, 77].”
BackgroundSign, Attend and Tell: Spatial Attention for Sign Language Recognition
Sarhan, N., & Frintrop, S. (2021) · International Conference on Automatic Face and Gesture Recognition · doi:10.1109/fg52635.2021.9666939 · Cited by 6
SRI3D: Two‐stream inflated 3D ConvNet based on sparse regularization for action recognition
Yang, Z., An, G., Zhang, R., Zheng, Z., & Ruan, Q. (2022) · IET Image Processing · doi:10.1049/ipr2.12725 · Cited by 6
Optical Flow-Guided Cine MRI Segmentation With Learned Corrections
Ortiz-Gonzalez, A., Kobler, E., Simon, S., Bischoff, L. M., Nowak, S., Isaak, A., Block, W., Sprinkart, A. M., Attenberger, U. I., Luetkens, J. A., et al. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3325766 · Cited by 6
“The proximal map of the data term reads as (for details we refer the reader to [49, Proposition 3]): Then, the algorithm is given by Initially, (cid:56) 0 coincides with the zero flow.”
MethodInfluentialA Rotation-Invariant Regularization Term for Optical Flow Related Problems
Palomares, R. P., Haro, G., & Ballester, C. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-16814-2_20 · Cited by 6
“We introduce an auxiliar variable v representing the optical flow, as in [10], and we penalize its deviation from u.”
BackgroundMethodInfluentialLow-cost sensor to detect overtaking based on optical flow
Guzmán, P., Díaz, J., Ralli, J., Agís, R., & Ros, E. (2011) · Machine Vision and Applications · doi:10.1007/s00138-011-0392-2 · Cited by 6
Optical flow A combined local-global approach using L1 norm
Drulea, M., Peter, I. R., & Nedevschi, S. (2010) · IEEE International Conference on Intelligent Computer Communication and Processing (ICCP) · doi:10.1109/iccp.2010.5606437 · Cited by 6
Illumination invariance for driving scene optical flow using comparagram preselection
Dederscheck, D., Muller, T., & Mester, R. (2012) · IEEE Intelligent Vehicles Symposium · doi:10.1109/ivs.2012.6232281 · Cited by 6
“L 1 optical flow estimation method by Zach and Pock [2], implemented by [18].”
MethodDifferential Optical Flow Estimation Under Monocular Epipolar Line Constraint
Mohamed, M. A., Mirabdollah, M. H., & Mertsching, B. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-20904-3_32 · Cited by 6
Motion-Adaptive Depth Superresolution
Kamilov, U. S., & Boufounos, P. T. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2658944 · Cited by 6
Dynamic turbulence mitigation with large moving objects
Nieuwenhuizen, R. P. J., van Eekeren, A. W. M., Dijk, J., & Klamer Schutte, B. (2017) · Electro-Optical and Infrared Systems: Technology and Applications · doi:10.1117/12.2277840 · Cited by 6
Robust Multi-Modal Cues for Dyadic Human Interaction Recognition
Trabelsi, R., Varadarajan, J., Pei, Y., Zhang, L., Jabri, I., Bouallègue, A., & Moulin, P. (2017) · doi:10.1145/3132515.3132517 · Cited by 6
“As for the dense optical ow computation from depth and RGB images, we used the TVL1 algorithm [26].”
MethodJoint Denoising and Demosaicking of Raw Video Sequences
Buades, A., & Duran, J. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451853 · Cited by 6
“We use the TV-L1 optical flow method introduced in [21].”
MethodHuman Action Recognition Based on a Spatio-Temporal Video Autoencoder
Santos, A. C. S. E., & Pedrini, H. (2019) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001420400017 · Cited by 6
Removing multi-frame Gaussian noise by combining patch-based filters with optical flow
Bodduna, K., & Weickert, J. (2021) · Journal of Electronic Imaging · doi:10.1117/1.jei.30.3.033031 · Cited by 6
“These motion estimation techniques perform better than some classical strategies [48], [49].”
MethodMulti-Task Learning of Generalizable Representations for Video Action Recognition
Yao, Z., Wang, Y., Long, M., Wang, J., Yu, P. S., & Sun, J. (2020) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme46284.2020.9102724 · Cited by 6
“,Oτ} are the target optical flow which are pre-computed using TLV1 [21].”
MethodBubblenet: A Disperse Recurrent Structure To Recognize Activities
Bastos, I. L. O., C., M. V. H., & Schwartz, W. R. (2020) · Proceedings - International Conference on Image Processing · doi:10.1109/icip40778.2020.9190769 · Cited by 6
“In turn, the Optical Flow Input corresponds to the DTVL-1 [26] algorithm response presented to stacked 2D convolutional layers, as shown in Figure 2.”
MethodYong, X., Huang, C.-K., & Lim, C. T. (2021) · Journal of The Royal Society Interface · doi:10.1098/rsif.2021.0248 · Cited by 6
“…dense Pyramidal Lucas Kanade optical flow (PyrLK) [23]; (c) Farneback optical flow [24]; (d) Brox optical flow [25]; (e) dual TV-L1 optical flow [26]; (2) perform FlowMatch (optional), a frame-to-frame object matching algorithm that is available with the inclusion of labelled images which…”
MethodKrishnaswamy, D., Hareendranathan, A. R., Suwatanaviroj, T., Boulanger, P., Becher, H. H., Noga, M., & Punithakumar, K. (2021) · Cardiovascular Engineering and Technology · doi:10.1007/s13239-021-00547-6 · Cited by 6
“For the 4D temporal segmentation, the proposed method was compared to four other state of the art methods, two variants of the demons algorithm from Insight ToolKit (ITK) [40], optical flow from the Scikit-Image [41,42,43, 44], and optical flow from the OpenCV package [46,47].”
MethodInfluentialYousaf, K., Nawaz, T., & Habib, A. (2023) · Multimedia Tools and Applications · doi:10.1007/s11042-023-15774-3 · Cited by 9
“Two types of dense optical flow estimations are evaluated for the proposed network: (1) Gunnar-Farneback [21], and (2) dual TV-L1 [87].”
Incremental projection approach of regularization for inverse problems
Souopgui, I., Ngodock, H., Vidard, A., & Dimet, F. L. (2015) · Applied Mathematics & Optimization · doi:10.1007/s00245-015-9315-3 · Cited by 6
“Additional techniques to improve the solution of motion estimation problems include the discretization scheme such as Helmhotz decomposition [50], the use of L1-norm [45,51], and multi-scale strategies [1,3,6,8,9,15,22,27,31] and references therein.”
MethodRobinson, P. E., & Nel, A. (2016) · Journal of Electronic Imaging · doi:10.1117/1.jei.25.6.063010 · Cited by 6
“In this work, we investigated a modernized Horn–Schunck (HS) algorithm implemented by Sun et al.,37 Farneback’s algorithm,38 and the Dual TV-L1 algorithm by Zach et al.39 The latter two algorithms have GPUbased implementations available in OpenCV.”
MethodAction Recognition by Composite Deep Learning Architecture I3D-DenseLSTM
Joefrie, Y. Y., & Aono, M. (2019) · International Conference on Advanced Informatics: Concepts, Theory and Applications (ICAICTA) · doi:10.1109/icaicta.2019.8904245 · Cited by 5
“When calculating optical flow, we use OpenCV built-in Brox algorithm, proposed by [36], as this algorithm is quite fast in calculating optical flow compared to TV-L1 algorithm [37].”
MethodActivity Detection in Untrimmed Videos Using Chunk-based Classifiers
Gleason, J., Schwarcz, S. L., Ranjan, R., Castillo, C. D., Chen, J.-C., & Chellappa, R. (2020) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw50321.2020.9096912 · Cited by 6
“At the time of this writing, the most powerful open source video classification algorithm is the Inflated 3D Con-volution (I3D) architecture [7], we take the 64 frames that compose it and compute optical flow for each one using the TV-L1 optical flow algorithm [57].”
MethodAerial Implicit 3D Video Stabilization Using Epipolar Geometry Constraint
Huynh, L., Choi, J., & Medioni, G. (2014) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2014.600 · Cited by 6
“To take advantage of the small variance between successive frames in video, we use the TotalVariance L(1) optical flow [23] to get dense point correspondences between successive frames, which is implemented by minimizing an L(1) norm data term and a Total-Variance regularization term:”
MethodA phase-based framework for optical flow estimation on omnidirectional images
Alibouch, B., Radgui, A., Demonceaux, C., Rziza, M., & Aboutajdine, D. (2014) · Signal Image and Video Processing · doi:10.1007/s11760-014-0739-z · Cited by 6
Visentini-Scarzanella, M., Merrifield, R. D., Stoyanov, D. V., & Yang, G. (2010) · Lecture notes in computer science · doi:10.1007/978-3-642-15711-0_33 · Cited by 6
“The nature of the data term depends on the optimisation procedure adopted: faster techniques involve the use of differentiable data terms [5], whereas more sophisticated non-differentiable penalty functions are coupled with complete searches [6].”
MethodMotion and acceleration from image assimilation with evolution models
Béréziat, D., & Herlin, I. (2018) · Digital Signal Processing · doi:10.1016/j.dsp.2018.08.008 · Cited by 6
Visual SLAM and Surface Reconstruction for Abdominal Minimally Invasive Surgery
Lin, B. (2015) · Digital Commons - University of South Florida (University of South Florida) · Cited by 6
A Flexible Primal-Dual ToolboxPreprint
Dirks, H. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.05835 · Cited by 6
“This formulation is non-linear in terms of v and therefore linearized (see e.g. [16, 4]).”
BackgroundConvolutional Relational Machine for Group Activity RecognitionPreprint
Azar, S. M., Atigh, M. G., Nickabadi, A., & Alahi, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.03308 · Cited by 6
“We use TVL1 algorithm [28] to compute these optical flow fields.”
MethodGeneral Dynamic Scene Reconstruction from Multiple View VideoPreprint
Mustafa, A., Kim, H., Guillemaut, J., & Hilton, A. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1509.09294 · Cited by 6
“Dynamic elements of the scene are identified by performing optical flow [37] on consecutive frames for a single view of each cluster.”
BackgroundTorpey, D., & Çelik, T. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.09423 · Cited by 6
“Furthermore, different, more accurate, optical flow algorithms could be used in place of Farneback’s methods (such as TVL-1 [29] or Brox [30]).”
MethodVideo Action Understanding: A Tutorial.Preprint
Hutchinson, M., & Gadepally, V. N. (2020) · arXiv (Cornell University) · Cited by 6
Optimization Planning for 3D ConvNetsPreprint
Qiu, Z., Yao, T., Ngo, C., & Mei, T. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2201.04021 · Cited by 6
“For fair comparisons with the two-stream baselines on Kinetics datasets, we additionally consider the flow modality by the two-direction optical flow image extracted by TV-L1 algorithm (Zach et al., 2007) and directly use the optimal path on RGB modality as that on flow modality.”
MethodVideo Classification with FineCoarse Networks.Preprint
Huang, G., & Borş, A. G. (2021) · arXiv (Cornell University) · Cited by 6
Gan, Y. S., Liong, G.-B., Liu, K., & Liong, S. (2023) · Neurocomputing · doi:10.1016/j.neucom.2023.03.055 · Cited by 5
Information theory-guided heuristic progressive multi-view coding
Li, J., Gao, H., Qiang, W., & Zheng, C. (2023) · Neural Networks · doi:10.1016/j.neunet.2023.08.027 · Cited by 5
Learning Temporal Dynamics in Videos With Image Transformer
Shu, Y., Qiu, Z., Long, F., Yao, T., Ngo, C., & Mei, T. (2024) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2024.3383662 · Cited by 7
“One typical way is to firstly calculate the pixel-level moving direction and velocity by dense optical flow [1], [2], and then feed the normalized optical flow as images into classification networks [3], [4], [5], [6], [7], [8].”
MethodParticle Tracking in Biological Images with Optical-Flow Enhanced Kalman Filtering
Reme, R., Newson, A., Angelini, E. D., Olivo‐Marín, J., & Lagache, T. (2024) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi56570.2024.10635656 · Cited by 5
“We have compared Farneback [19], TVL1 [20, 21] and deep-learning based RAFT [22] algorithms on a video of Hydra Vulgaris neurons [23].”
MethodCPR-Coach: Recognizing Composite Error Actions Based on Single-Class Training
Wang, S., Wang, S., Yang, D., Li, M., Kuang, H., Zhao, X. Y., Su, L., Zhai, P., & Zhang, L. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.01777 · Cited by 5
“The CPR-Coach also provides optical flow images generated by the TV-L1 algorithm [66] and 2D skeletons of the rescuer obtained by Alphapose [20].”
MethodLu, J., He, R., Zhou, S., Tan, W., & Yan, B. (2024) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3664647.3680921 · Cited by 6
“Initially approached as an energy minimization problem, it utilized human-designed data and prior terms as optimization objectives [3–6 , 51].”
MethodQi, Y., Ibrayim, M., & Hamdulla, A. (2024) · IEEE International Joint Conference on Biometrics (IJCB) · doi:10.1109/ijcb62174.2024.10744439 · Cited by 5
Singh, B., Kumar, P., & Jain, S. K. (2025) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-024-01352-y · Cited by 5
“Therefore, the convex relaxation term is added such that ( u , v ) and ( ¯ u , ¯ v ) are in close approximation [46].”
Estimating full-field displacement in biological images using deep learning
Warsop, S. J. E. T., Caixeiro, S., Bischoff, M., Kursawe, J., Bruce, G. D., & Wijesinghe, P. (2025) · npj Artificial Intelligence · doi:10.1038/s44387-025-00005-x · Cited by 5
“For this work, we used an implementation of the LV-L1 method [52] in an image pyramid form provided by scikit-image [53].”
Micro-Expression Recognition via Fine-Grained Dynamic Perception
Shao, Z., Cheng, Y., Zhang, F., Shi, X., Li, C., Ma, L., & Yeung, D. (2025) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3765901 · Cited by 5
“Moreover, these features like optical flow [72] often additionally rely on key frames including onset, apex, and offset frames of MEs to improve the recognition performance, which limits the applicability.”
Optical Flow in Driver Assistance Systems
Onkarappa, N. (2014) · ELCVIA Electronic Letters on Computer Vision and Image Analysis · doi:10.5565/rev/elcvia.593 · Cited by 5
“TV-L1 technique is a variational optical flow technique proposed in [107] that gives dense flow field.”
BackgroundMethodInfluentialA loop-consistency measure for dense correspondences in multi-view video
Sellent, A., Ruhl, K., & Magnor, M. A. (2012) · Image and Vision Computing · doi:10.1016/j.imavis.2012.06.011 · Cited by 5
“We apply an image pyramid for optical flow estimations [39, 38, 21], see Alg.”
BackgroundMethodAn improved periodic optical flow model for cardiac gated image reconstruction
Qi, W., Niu, X., & Yang, Y. (2011) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi.2011.5872634 · Cited by 5
“Based on this prior success, in this work we further develop this temporal modeling approach by exploring the use of a smoothness term based on the total variation (TV) of the motion field [12], which potentially can better accommodate the discontinuities in the motion field associated with an object boundary.”
MethodActivity Detection and Recognition of Daily Living Events
Avgerinakis, K., Briassouli, A., & Kompatsiaris, I. Y. (2015) · Springer eBooks · doi:10.1007/978-3-319-17963-6_8 · Cited by 5
Second-order motion descriptors for efficient action recognition
García, R. O., Morales, E. F., & Sucar, L. E. (2020) · Pattern Analysis and Applications · doi:10.1007/s10044-020-00924-2 · Cited by 5
“Zach et al. [10] utilize the RANSAC algorithm to reduce the motion induced by the camera, thus becoming a powerful but heavy algorithm.”
MethodInfluentialRenders, J., Sijbers, J., & Beenhouwer, J. D. (2021) · Medical Physics · doi:10.1002/mp.14765 · Cited by 5
Seznec, M., Nicolas Gac, A., Orieux, F., & Naik, A. S. (2021) · Journal of Real-Time Image Processing · doi:10.1007/s11554-021-01187-8 · Cited by 5
“Brox et al. add a gradient conservation term [3] while Zach et al. use a L1-norm penalization instead of a quadratic one [26] to obtain better-defined object boundaries.”
BackgroundDevelopment of an IR Video Surveillance System Based on Fractional Order TV-Model
Kumar, P., Khan, M., & Gupta, S. (2021) · International Conference on Control, Automation, Power and Signal Processing (CAPS) · doi:10.1109/caps52117.2021.9730605 · Cited by 5
Weakly-Supervised Temporal Action Localization with Multi-Head Cross-Modal Attention
Ren, H., Ren, H., Ran, W., Lu, H., & Jin, C. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-20868-3_21 · Cited by 5
Wang, Y., Wang, Y., Duan, W., Zheng, Y., & Duan, P. (2023) · Journal of Circuits Systems and Computers · doi:10.1142/s0218126624500518 · Cited by 5
Ballot Tabulation Using Deep Learning
Zhao, F., Zhang, C., Saxena, N., Wallach, D. S., & Rabby, A. S. A. (2023) · IEEE International Conference on Information Reuse and Integration (IRI) · doi:10.1109/iri58017.2023.00026 · Cited by 5
“Therefore, an optical-flow-based fine-level alignment method is adopted [12–14].”
MethodXia, L., & Fu, W. (2024) · Cluster Computing · doi:10.1007/s10586-024-04553-w · Cited by 5
“We extract the optical flow using the TVL1 optical flow algorithm [42] and use a stack of 10 continuous optical flows as input to the optical stream network.”
Mild cognitive impairment prediction based on multi-stream convolutional neural networks
Lee, C.-C., Chau, H.-H. H., Wang, H.-L., Chuang, Y., & Chau, Y. A. (2024) · BMC Bioinformatics · doi:10.1186/s12859-024-05911-6 · Cited by 7
“In this study, we choose the TVL1 optical flow algorithm [37] implemented by OpenCV with CUDA.”
Method
Metadata
- Authors
- Zach, C., Pock, T., & Bischof, H.
- Affiliations
- VRVis GmbH (Austria) (AT)
- Graz University of Technology (AT)
- Venue
- Lecture Notes in Computer Science
- Publisher
- Springer Berlin Heidelberg
- Published
- 2007-09-21
- Pages
- 214-223
- OpenAlex
- W1578985305
- Cited by
- 1,658
- References
- 27
- Citations by year
- 2026: 44 · 2025: 62 · 2024: 83 · 2023: 111 · 2022: 107 · 2021: 188 · 2020: 152 · 2019: 149 · 2018: 112 · 2017: 98 · 2016: 88 · 2015: 99 · 2014: 70 · 2013: 59 · 2012: 56
- FWCI
- 27.93
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Springer eBooks (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Vision and Imaging
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Video Surveillance and Tracking Methods
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Signal Denoising Methods
- Keywords
- optical flow, total variation regularization, GPU acceleration, real-time video processing, variational methods, discontinuity preservation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
