Fundamentals of Computing Continuous Dynamic Time Warping in 2D under Different Norms
Pith reviewed 2026-05-17 04:47 UTC · model grok-4.3
The pith
Continuous dynamic time warping cannot be computed exactly under the Euclidean 2-norm using only algebraic operations.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We show that CDTW cannot be computed exactly under the Euclidean 2-norm using only algebraic operations, and we give an exact algorithm for CDTW under norms approximating the 2-norm. The latter result relies on technical fundamentals that we establish, and which generalise to any norm and to related measures such as the partial Fréchet similarity.
What carries the argument
Technical fundamentals for exact CDTW computation under norms that approximate the Euclidean 2-norm, enabling both the algorithm and its generalization to arbitrary norms.
If this is right
- Exact CDTW algorithms become available for every norm once the fundamentals are applied.
- The same fundamentals yield exact algorithms for partial Fréchet similarity.
- Designers of curve-similarity tools can avoid the algebraic impossibility by switching to a suitable approximating norm.
Where Pith is reading between the lines
- In practice, Euclidean CDTW may require numerical or approximate methods rather than closed-form algebraic solutions.
- The non-computability result could guide similar investigations for other robust curve distances.
- Applications in motion analysis or pattern recognition gain concrete guidance on when exact solutions are feasible.
Load-bearing premise
The technical fundamentals established for approximating norms are assumed to generalize without additional restrictions to any norm and to related measures such as partial Fréchet similarity.
What would settle it
An explicit algebraic formula that computes exact CDTW under the Euclidean 2-norm, or a concrete norm or measure for which the established fundamentals fail to produce an exact algorithm.
Figures
read the original abstract
Continuous Dynamic Time Warping (CDTW) measures the similarity of polygonal curves robustly to outliers and to sampling rates, but the design and analysis of CDTW algorithms face multiple challenges. We show that CDTW cannot be computed exactly under the Euclidean 2-norm using only algebraic operations, and we give an exact algorithm for CDTW under norms approximating the 2-norm. The latter result relies on technical fundamentals that we establish, and which generalise to any norm and to related measures such as the partial Fr\'echet similarity.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims that Continuous Dynamic Time Warping (CDTW) for polygonal curves cannot be computed exactly under the Euclidean 2-norm using only algebraic operations. It presents an exact algorithm for CDTW under norms that approximate the 2-norm, based on newly established technical fundamentals. These fundamentals are asserted to generalize to arbitrary norms and to related measures such as partial Fréchet similarity.
Significance. If the impossibility result and the exact algorithm for approximating norms are correct, the work clarifies fundamental computational limits of CDTW and supplies a practical exact method via norm approximation. The claimed generalization of the technical fundamentals would broaden the results' impact to general norms and measures like partial Fréchet similarity, supporting more robust curve-similarity algorithms in computational geometry that tolerate sampling-rate variation and outliers.
major comments (1)
- Abstract and the section establishing the technical fundamentals: the assertion that these fundamentals 'generalise to any norm and to related measures such as the partial Fréchet similarity' is stated without an explicit extension theorem or derivation. If the proofs for approximating norms rely on properties such as strict convexity of the unit ball or differentiability away from axes, the generalization step is load-bearing for the broader utility claim and requires a concrete argument showing the properties transfer without additional restrictions.
minor comments (2)
- Clarify the precise definition of 'norms approximating the 2-norm' (e.g., via a specific distance or convergence criterion) at the first point of use.
- Add a short discussion or reference to prior work on algebraic computation of Fréchet distance to contextualize the impossibility result for CDTW.
Simulated Author's Rebuttal
We thank the referee for their careful and constructive review of our manuscript. The feedback on the presentation of the generalization claim is helpful, and we address it directly below.
read point-by-point responses
-
Referee: Abstract and the section establishing the technical fundamentals: the assertion that these fundamentals 'generalise to any norm and to related measures such as the partial Fréchet similarity' is stated without an explicit extension theorem or derivation. If the proofs for approximating norms rely on properties such as strict convexity of the unit ball or differentiability away from axes, the generalization step is load-bearing for the broader utility claim and requires a concrete argument showing the properties transfer without additional restrictions.
Authors: We appreciate the referee highlighting the need for greater explicitness on this point. The technical fundamentals are developed from the combinatorial structure of the free-space diagram and the monotonicity properties of admissible warping paths; these rely only on the axioms of a norm (positivity, homogeneity, and the triangle inequality) together with the polygonal nature of the input curves. No appeal is made to strict convexity of the unit ball. Differentiability away from the axes is used solely to guarantee that the critical points arising under the approximating norms remain algebraic; the same structural lemmas carry over to an arbitrary norm by direct substitution of the norm into the distance function, with the algorithm adapting via the same sequence of candidate intervals. The extension to partial Fréchet similarity follows identically by restricting the admissible regions to sub-curves while preserving the same monotonicity and continuity arguments. We will revise the manuscript by inserting a short dedicated subsection (or appendix) that states the extension as a formal corollary and sketches the transfer of each key lemma, thereby making the generalization fully explicit without imposing further restrictions. revision: yes
Circularity Check
No circularity: derivation chain is self-contained with new technical fundamentals
full rationale
The paper establishes an impossibility result for exact algebraic CDTW under the Euclidean 2-norm and provides an exact algorithm for approximating norms, relying on technical fundamentals developed within the work. These fundamentals are presented as newly derived and asserted to generalize, without any quoted reduction of outputs to fitted parameters, self-definitions, or load-bearing self-citations that collapse the central claims back to inputs by construction. No equations or sections exhibit patterns such as renaming a fit as a prediction or importing uniqueness via prior author work. The derivation remains independent and self-contained against external benchmarks.
Axiom & Free-Parameter Ledger
axioms (1)
- standard math Polygonal curves are piecewise-linear and norms satisfy standard metric properties.
Lean theorems connected to this paper
-
IndisputableMonolith/Foundation/AlexanderDuality.leanalexander_duality_circle_linking unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
We show that CDTW cannot be computed exactly under the Euclidean 2-norm using only algebraic operations, and we give an exact algorithm for CDTW under norms approximating the 2-norm. The latter result relies on technical fundamentals that we establish, and which generalise to any norm and to related measures such as the partial Fréchet similarity.
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
Theorem 8. ... There exists a valley ℓ of positive slope for P, Q under ∥·∥.
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Forward citations
Cited by 1 Pith paper
-
A Constant-Factor Approximation for Continuous Dynamic Time Warping in 2D
A 5-approximation algorithm for 2D continuous dynamic time warping under the 1-norm with O(n^5) time, extendable to (5+ε) for any fixed norm.
Reference graph
Works this paper leans on
-
[1]
Helmut Alt and Michael Godau. Computing the Fréchet distance between two polygonal curves.International Journal of Computational Geometry & Applications, 5(1–2):75–91, 1995.doi:10.1142/S0218195995000064. 14 K. Buchin, M. Buchin, J. E. Swiadek, S. Wong
-
[2]
ChandrajitBajaj. Thealgebraicdegreeofgeometricoptimizationproblems.Discrete & Computational Geometry, 3(2):177–191, 1988.doi:10.1007/BF02187906
-
[3]
Cambridge University Press, 1990
Alan Baker.Transcendental Number Theory. Cambridge University Press, 1990. Reissue with updated material.doi:10.1017/CBO9780511565977
-
[4]
Cambridge Univer- sity Press, 7th edition, 2009
Stephen Boyd and Lieven Vandenberghe.Convex Optimization. Cambridge Univer- sity Press, 7th edition, 2009. URL:https://web.stanford.edu/~boyd/cvxbook/
work page 2009
-
[5]
On map- matching vehicle tracking data
Sotiris Brakatsoulas, Dieter Pfoser, Randall Salas, and Carola Wenk. On map- matching vehicle tracking data. InProceedings of the 31st International Conference on Very Large Data Bases, pages 853–864. VLDB Endowment, 2005. URL:https: //dl.acm.org/doi/10.5555/1083592.1083691
-
[6]
PhD thesis, University of Sydney, 2022
Milutin Brankovic.Graphs and Trajectories in Practical Geometric Problems. PhD thesis, University of Sydney, 2022
work page 2022
-
[7]
Milutin Brankovic, Kevin Buchin, Koen Klaren, André Nusser, Aleksandr Popov, and Sampson Wong.( k, l)-medians clustering of trajectories using Continuous Dynamic Time Warping. InProceedings of the 28th International Conference on Advances in Geographic Information Systems, pages 99–110. ACM, 2020. doi: 10.1145/3397536.3422245
-
[8]
Kevin Buchin, Maike Buchin, Wouter Meulemans, and Wolfgang Mulzer. Four soviets walk the dog: Improved bounds for computing the Fréchet distance.Discrete & Computational Geometry, 58(1):180–216, 2017.doi:10.1007/S00454-017-987 8-7
-
[9]
Exact algorithms for partial curve matching via the Fréchet distance
Kevin Buchin, Maike Buchin, and Yusu Wang. Exact algorithms for partial curve matching via the Fréchet distance. InProceedings of the Twentieth Annual ACM- SIAM Symposium on Discrete Algorithms, pages 645–654. SIAM, 2009. doi: 10.1137/1.9781611973068.71
-
[10]
Computing Continuous Dynamic Time Warping of time series in polynomial time
Kevin Buchin, André Nusser, and Sampson Wong. Computing Continuous Dynamic Time Warping of time series in polynomial time. In38th International Symposium on Computational Geometry, pages 22:1–22:16. Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2022.arXiv:2203.04531,doi:10.4230/LIPICS.SOCG.2022.22
-
[11]
PhD thesis, Freie Universität Berlin, 2007
Maike Buchin.On the Computability of the Fréchet Distance between Triangulated Surfaces. PhD thesis, Freie Universität Berlin, 2007. URL:https://refubium.f u-berlin.de/handle/fub188/1909
work page 2007
-
[12]
Bernard Chazelle and David P. Dobkin. Intersection of convex objects in two and three dimensions.Journal of the ACM, 34(1):1–27, 1987.doi:10.1145/7531.24036
-
[13]
Fréchet distance in subquadratic time
Siu-Wing Cheng and Haoqiang Huang. Fréchet distance in subquadratic time. In Proceedings of the 2025 Annual ACM-SIAM Symposium on Discrete Algorithms, pages 5100–5113. SIAM, 2025.doi:10.1137/1.9781611978322.173
-
[14]
Similarity of polygonal curves in the presence of outliers
Jean-Lou De Carufel, Amin Gheibi, Anil Maheshwari, Jörg-Rüdiger Sack, and Christian Scheffer. Similarity of polygonal curves in the presence of outliers. Computational Geometry, 47(5):625–641, 2014.doi:10.1016/J.COMGEO.2014.01.0 02
-
[15]
A note on the unsolvability of the weighted region shortest path problem
Jean-Lou De Carufel, Carsten Grimm, Anil Maheshwari, Megan Owen, and Michiel Smid. A note on the unsolvability of the weighted region shortest path problem. Computational Geometry, 47(7):724–727, 2014.doi:10.1016/J.COMGEO.2014.02.0 04
-
[16]
Alon Efrat, Quanfu Fan, and Suresh Venkatasubramanian. Curve matching, time warping, and light fields: New algorithms for computing similarity between curves. Journal of Mathematical Imaging and Vision, 27(3):203–216, 2007. doi:10.1007/ S10851-006-0647-0
work page 2007
-
[17]
Maurice Fréchet. Sur quelques points du calcul fonctionnel.Rendiconti del Circolo Matematico di Palermo, 22(1):1–72, 1906.doi:10.1007/BF03018603. Fundamentals of Computing CDTW in 2D under Different Norms 15
-
[18]
Omer Gold and Micha Sharir. Dynamic Time Warping and Geometric Edit Distance: Breaking the quadratic barrier.ACM Transactions on Algorithms, 14(4):50:1–50:17, 2018.doi:10.1145/3230734
-
[19]
Sariel Har-Peled, Benjamin Raichel, and Eliot W. Robson. The Fréchet distance unleashed: Approximating a dog with a frog. In41st International Symposium on Computational Geometry, pages 54:1–54:13. Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2025.arXiv:2407.03101,doi:10.4230/LIPICS.SOCG.2025.54
-
[20]
Alan Jeffrey and Hui-Hui Dai.Handbook of Mathematical Formulas and Integrals. Elsevier, 4th edition, 2008
work page 2008
-
[21]
Continuous Dynamic Time Warping for clustering curves
Koen Klaren. Continuous Dynamic Time Warping for clustering curves. Master’s thesis, Eindhoven University of Technology, 2020. URL:https://research.tue.n l/en/studentTheses/continuous-dynamic-time-warping-for-clustering-cur ves
work page 2020
-
[22]
Wilhelmus A. J. Luxemburg. Arzelà’s Dominated Convergence Theorem for the Riemann integral.The American Mathematical Monthly, 78(9):970–979, 1971. doi:10.1080/00029890.1971.11992915
-
[23]
Approximating the integral Fréchet distance.Computational Geometry, 70–71:13–30, 2018
Anil Maheshwari, Jörg-Rüdiger Sack, and Christian Scheffer. Approximating the integral Fréchet distance.Computational Geometry, 70–71:13–30, 2018. doi: 10.1016/J.COMGEO.2018.01.001
-
[24]
Mario E. Munich and Pietro Perona. Continuous Dynamic Time Warping for translation-invariant curve alignment with applications to signature verification. In Proceedings of the Seventh IEEE International Conference on Computer Vision, pages 108–115. IEEE, 1999.doi:10.1109/ICCV.1999.791205
-
[25]
Lexicographic Fréchet matchings
Günter Rote. Lexicographic Fréchet matchings. Extended abstract from the 30th European Workshop on Computational Geometry, 2014. URL:https://refubium .fu-berlin.de/handle/fub188/15658
work page 2014
-
[26]
Helmut H. Schaefer and Manfred P. Wolff.Topological Vector Spaces. Springer, 2nd edition, 1999.doi:10.1007/978-1-4612-1468-7
-
[27]
Subpixel contour matching using continuous dynamic programming
Bruno Serra and Marc Berthod. Subpixel contour matching using continuous dynamic programming. In1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 202–207. IEEE, 1994.doi:10.1109/CVPR.1 994.323830
-
[28]
Optimal subpixel matching of contour chains and segments
Bruno Serra and Marc Berthod. Optimal subpixel matching of contour chains and segments. InProceedings of IEEE International Conference on Computer Vision, pages 402–407. IEEE, 1995.doi:10.1109/ICCV.1995.466911
-
[29]
Han Su, Shuncheng Liu, Bolong Zheng, Xiaofang Zhou, and Kai Zheng. A survey of trajectory distance measures and performance evaluation.The VLDB Journal, 29(1):3–32, 2020.doi:10.1007/S00778-019-00574-9
-
[30]
GIScience & Remote Sensing59(1), 200–214 (2022) https: //doi.org/10.1080/15481603.2021.2023840
Yaguang Tao, Alan Both, Rodrigo I. Silveira, Kevin Buchin, Stef Sijben, Ross S. Purves, Patrick Laube, Dongliang Peng, Kevin Toohey, and Matt Duckham. A comparative analysis of trajectory similarity measures.GIScience & Remote Sensing, 58(5):643–669, 2021.doi:10.1080/15481603.2021.1908927
-
[31]
Taras K. Vintsyuk. Speech discrimination by dynamic programming.Cybernetics and Systems Analysis, 4(1):52–57, 1968.doi:10.1007/BF01074755
-
[32]
Norman J. Wildberger and Dean Rubine. A hyper-Catalan series solution to polynomial equations, and the Geode.The American Mathematical Monthly, 132(5):383–402, 2025.doi:10.1080/00029890.2025.2460966. 16 K. Buchin, M. Buchin, J. E. Swiadek, S. Wong A Additional Proofs for Section 2 For the sake of completeness, we give proofs of Theorem 5 and Lemma 6. Theo...
-
[33]
+ 7/2)2 = 99/8 + 7/(2· √ 2)>100/8 = 25/2 = (5/ √ 2)2.⊓ ⊔ So far, we have avoided to evaluate the integrals via antiderivatives, but the third claim relies on this to obtainC′ as a linear combination of logarithms, which allows to apply transcendental number theory by means of Lemma 10. Claim (B3).Some algebraic functions α1, α2, β0, β1, β2 : [0, 10] \ {2,...
-
[34]
= 0is fulfilled on(0 , 10)only by s∗ 0 := 1/929 · 1880· √ 10−3070 + 60· p 16· √ 10 + 61 / 80· √ 10 + 269 ∈(4,4.5). Proof. Let h: R→R be a function withh(s) := √s2 +λ 1s+λ 0 for λ0, λ1 ∈R . By [20, Equation 4.3.4.1.2], the antiderivative oft7→t/h(t)is given by Z t h(t) dt=− λ1 2 ·ln(|2·h(s) + 2s+λ 1|) +h(s) +C. Fundamentals of Computing CDTW in 2D under Di...
-
[35]
L. Paul Chew and Robert L. (Scot) Drysdale, III. Voronoi diagrams based on convex distance functions. InProceedings of the First Annual Symposium on Computational Geometry, pages 235–244. ACM, 1985.doi:10.1145/323233.323264
-
[36]
Mark de Berg, Otfried Cheong, Marc van Kreveld, and Mark Overmars.Com- putational Geometry: Algorithms and Applications. Springer, 3rd edition, 2008. doi:10.1007/978-3-540-77974-2
-
[37]
John Hershberger. Finding the upper envelope ofn line segments in O(nlogn ) time.Information Processing Letters, 33(4):169–174, 1989. doi:10.1016/0020-0 190(89)90136-1
-
[38]
Murray H. Protter and Charles B. Morrey, Jr.Intermediate Calculus. Springer, 2nd edition, 1985.doi:10.1007/978-1-4612-1086-3
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.