{"id":"995366bd-c026-496b-9bbb-d46abf62e0fe","arxiv_id":"2412.09200","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"New asymptotic blends and Taylor extrapolations improve two families of distance-to-boundary estimators, demonstrated on 2D images.","lead":"This paper makes two accuracy upgrades for estimating distance to a shape's boundary: it blends two convolutional estimates using Laplace asymptotics, and it adds Taylor corrections to the heat method. Tests on 2D images show lower errors, though the main new scheme lacks a rigorous error bound.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Taylor extrapolation in s=1/λ behind (17)/(19) has no remainder bound and is tested far from the s→0 regime; the claimed accuracy gain over the heat method is therefore not established.","rationale":"The reader's weakest assumption identifies exactly the unproven smoothness and remainder control for the Taylor extrapolation used to derive (17) and (19). This is the most load-bearing concern because the central practical claim is that normalized (19) improves accuracy over the heat method (12), and that claim depends on the second-order Taylor extrapolation being accurate in the tested parameter range. The formal manipulations in Section IV are internally consistent: (15) and (17) coincide, and (19) is the correct second-order analogue. The paper's own Section VI admits the absence of rigorous justification, which corroborates the concern. The numerical evidence is honest but limited to a few shapes, without released code or quantitative error tables, so it does not settle the remainder question. The proposed unit-disk test is a clean analytical check because the screened-Poisson solution is known in closed form and the exact distance is known, so discretization error is not a confound. If that test shows the remainder is small and (19) indeed beats (12), the main concern is resolved; otherwise the conditional verdict should be tightened. Since the reader already chose CONDITIONAL and the concern is the same one that motivated that verdict, no verdict change is needed: the appropriate next step is the proposed test.","tokens_in":6488,"tokens_out":13095,"duration_ms":134485,"concrete_test":"For the unit disk of radius R, use the exact screened-Poisson solution v(r)=I0(λr)/I0(λR) to compute u=−λ⁻¹log v, u'_λ, and u''_λλ analytically. For t=1/λ² in {0.2,1,5,10} and r/R in {0.1,...,0.9}, compute the true Taylor remainder R2 = dist − [u − s u_s + (s²/2)u_ss] with s=1/λ, and compare it with the errors of (12), (15), and (19), including normalized versions. If |R2| is not small relative to the error of (12) at the recommended t, or if (19) does not beat (12) on the disk, then the reported improvement is not attributable to the Taylor extrapolation and the general claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the Taylor extrapolation (17)/(18). To conclude dist = u(0) ≈ u(s) − s u_s(s) + (s²/2)u_ss(s) with controlled error, one needs u(·) to be smooth in s near 0 and the remainder R(s) to be o(s²) uniformly on Ω. The paper never bounds R(s), and Section VI explicitly concedes that (15)/(19) are not supported by truly rigorous mathematical results. The concern is not merely formal: as s→0, u(s) converges to the distance function, which is only Lipschitz at the medial axis, so correction terms and their derivatives can be singular there, and for non-smooth domains non-analytic terms in s may enter the expansion. The numerical validation uses a small set of binary shapes, and the recommended parameter values (e.g., t=5, so s≈2.24) are far from the asymptotic limit s→0. Thus the observed superiority of normalized (19) over (12) is an empirical observation on a narrow test set, not a consequence of the demonstrated Taylor argument. If the remainder is not negligible in this parameter range on an independent reference domain, the central accuracy claim fails.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper addresses the problem of approximating the distance to the boundary of a bounded domain. It proposes two types of improvements. First, for convolutional distance transforms, it uses Laplace's method to derive a blend (Eq. (5)) of the log-sum-exp approximation (6) and the soft-min approximation (7), with a free constant K. Second, for differential (heat-method-type) approximations, it derives first- and second-order corrections (15) and (19) by differentiating the screened Poisson equation (9) with respect to λ and interpreting the result as a Taylor extrapolation in s=1/λ of the solution to the rescaled equation (11). The corrected schemes are then combined with the standard gradient-normalization step of the heat method. Numerical experiments on 2D binary shapes suggest that the normalized second-order scheme (19) outperforms the standard heat method (12). The paper is clearly written, and the algebraic steps from (9) through (19) are internally consistent.","tokens_in":6758,"tokens_out":5049,"duration_ms":50130,"significance":"If the accuracy gains were rigorously established, the paper would make a useful contribution to distance estimation: the proposed modifications are simple, require only one or two additional Poisson solves, and could improve the accuracy of heat-method-based distance computations in applications. The Laplace-method perspective connecting convolutional distance transforms to asymptotic expansions is also instructive. However, the central theoretical claim rests on a Taylor extrapolation for which no remainder estimate is provided, and the numerical validation is limited to a few 2D shapes without error bars or sensitivity analysis. The paper cannot currently be regarded as establishing the claimed superiority of (19) over the standard heat method, although the empirical evidence, especially the error plots in Fig. 2, is suggestive.","major_comments":[{"comment":"The Taylor extrapolation underlying (15) and (19) is the load-bearing step of the paper, but it is not rigorously justified. Equations (17) and (18) assume that u(s), the solution of (11) with s=1/λ, is sufficiently smooth in s near 0 and that the remainder of the Taylor expansion at s=1/λ is negligible uniformly on Ω. No bound on the remainder is provided, and Section VI explicitly admits that these schemes 'are not supported by truly rigorous mathematical results.' This is not a purely formal concern: as s→0, u(s) converges to the distance function, which is only Lipschitz at the medial axis, so the correction terms u'_s and u''_ss can be singular there. Furthermore, the parameter values used in the experiments (e.g., t=5, so s≈2.24) are far from the s→0 limit in which the asymptotic argument would apply. Thus the theoretical derivation does not, by itself, establish that (19) improves on (12). I recommend either providing a uniform remainder estimate under explicit regularity assumptions, or substantially expanding the numerical evidence so that the accuracy claim is backed by convergence tests as s→0 and by experiments on domains with medial-axis singularities.","section":"Section VI and Eqs. (17)-(19)"},{"comment":"The numerical evaluation of the central claim is too narrow. The quantitative error plots in Fig. 2 are for a single 2D shape, and the error maps in Fig. 3 are presented without numerical error values or comparisons across a dataset. There are no error bars and no statistical summary over multiple shapes or boundary discretizations, so the statement that the normalized version of (19) 'demonstrates the best performance' is not supported with the usual precision expected of a numerical paper. Additionally, the paper does not compare the proposed schemes against other recent distance estimators (e.g., the ADMM-based scheme of [14], which is from the same authors) beyond the heat method. Since the main claim is empirical accuracy improvement, the authors should provide a more systematic experimental study, including error statistics for a set of reference shapes and a convergence study with respect to the discretization.","section":"Section V, Figs. 2 and 3"},{"comment":"The convolutional blending formula (5) depends on a constant K that is not determined analytically. The paper sets K=0.1 heuristically and reports only one experiment for this value. Because the weights α and β in (5) are monotone functions of K, the observed accuracy improvement of the blend over the individual approximations (6) and (7) could depend critically on this free parameter. A sensitivity analysis over a range of K values, or an adaptive way of estimating K, is needed to establish the robustness of the convolutional improvement. As written, the experimental support for this contribution is a single shape with K fixed, which is insufficient.","section":"Section III, Eq. (5)"}],"minor_comments":[{"comment":"The definitions of α and β in (5) use K, but K is introduced only as an unspecified positive constant in (4); the text later states that K is heuristic, but this dual role should be made explicit in the notation.","section":"Section II, Eq. (5)"},{"comment":"The notation v'_λ and v''_λλ is used to denote derivatives with respect to λ; please state this explicitly at first use, as the same notation could be confused with a spatial derivative.","section":"Section IV, paragraph after Eq. (13)"},{"comment":"There is a typo in 'dist(x∂Ω': the argument should be 'dist(x, ∂Ω)'.","section":"Section V, paragraph on gradient normalization"},{"comment":"The statement that the Matlab implementations 'will be made available upon acceptance' makes reproducibility conditional; please provide a permanent repository or include the code at submission time.","section":"Section VI, reproducibility statement"}],"recommendation":"major_revision","confidential_remarks":"The paper is potentially acceptable after major revision. The main technical gap—the absence of remainder bounds for the Taylor extrapolation—is acknowledged by the authors, and the numerical evidence is suggestive but not conclusive. If the authors can provide a rigorous error estimate under suitable regularity assumptions or at least a much more thorough numerical study with convergence checks and sensitivity analysis, the paper would be a useful contribution. Note also that the code is promised only upon acceptance, which may conflict with the journal's reproducibility expectations."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick read of arXiv:2412.09200. The paper does two genuinely new things: it blends LogConv and SoftMin distance estimates using Laplace asymptotics with weights (5), and it derives differential distance formulas (15) and (19) by differentiating the screened Poisson equation with respect to λ. The algebra linking the convolutional identity (8) to the differential formula (14) is clean, and the connection to Laplace's method in [10] is worth having. The authors are also admirably honest in Section VI that the Taylor extrapolation schemes lack rigorous mathematical justification.\n\nThat honesty matches the main soft spot. The load-bearing step is the Taylor extrapolation in s = 1/λ behind (17)/(18). The remainder is never bounded, and the stress-test note is right that this is not merely formal: the distance function is only Lipschitz at the medial axis, so singular behavior is plausible, and the recommended t = 5 gives s ≈ 2.24, far from the s → 0 limit. So the accuracy gain of normalized (19) over the heat method is an empirical observation on a handful of 2D binary images, not a consequence of the Taylor argument. The experiments themselves are not misleading: error curves are shown over 0.2 ≤ t ≤ 10, normalized versions are compared fairly against the normalized heat method, and the conclusions are phrased as performance on tested shapes. Still, no code is released (promised on acceptance), no error bars, no other domain types, so the generality is unproven.\n\nThe convolutional blending part is on firmer footing since it rests on established Laplace asymptotics, though the constant K is heuristic. That part is a modest but sensible incremental contribution.\n\nOverall, this is a compact, readable letter that does not oversell. A serious referee should engage with it—the formal manipulations are solid, the empirical claims are honestly scoped, and the main request should be for a remainder estimate or at least a broader benchmark with released code before the accuracy improvements are treated as established.\n\nMy recommendation: send it to review, not desk reject.","headline":"A clean, honest incremental letter: new differential Taylor formulas for distance estimation with a genuinely unproven error analysis, so the empirical gains are plausible but not established.","tokens_in":7258,"tokens_out":1497,"would_cite":false,"duration_ms":15558,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65D18"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper derives first- and second-order corrections to heat-method distance estimates by differentiating the screened Poisson equation with respect to its decay parameter, and reports that the normalized second-order formula is the most…","keywords":["distance function approximation","heat method","screened Poisson equation","Laplace's method","Taylor extrapolation","convolutional distance transform","eikonal equation","gradient normalization"],"falsifier":"Use a domain with an exactly known distance function, such as a disk or annulus, solve the screened Poisson equation and its $\\lambda$-derivative equations numerically for many values of $\\lambda$, and compare the three formulas with the exact distance; if the second-order formula does not consistently beat the first-order one as $\\lambda$ grows, or if the improvement disappears when the heuristic gradient-normalization step is removed, the central accuracy claim would be disproved.","tokens_in":6256,"feed_emoji":"📏","tokens_out":9753,"duration_ms":92613,"temperature":0.7,"pith_summary":"This paper is about a basic geometry question: given a region, estimate for each interior point how far it is from the boundary. It offers two accuracy upgrades. For convolutional distance transforms, it blends the log-sum-exp and soft-min approximations using weights derived from Laplace's method. For differential methods, it differentiates the screened Poisson equation with respect to the decay parameter and uses the first- and second-order Taylor terms as new distance estimates. The paper's numerical evidence on 2D binary shapes says the normalized second-order estimate is the most accurate, with a flat error curve that makes it less sensitive to the choice of parameter.","feed_headline":"A second-order Poisson formula sharpens distance estimates","feed_subtitle":"Differentiating the screened Poisson equation by its decay rate beats the standard heat-method accuracy on 2D shapes.","key_machinery":"The machinery is the $\\lambda$-derivative of the screened Poisson solution. Differentiating the boundary-integral representation with respect to $\\lambda$ turns the logarithm of the heat kernel into a new distance estimator, $-v'_\\lambda/v$, the continuous analogue of the identity that turns log-sum-exp into a soft minimum. Taylor expansion of $u(x;s)$ in $s = 1/\\lambda$ around $s = 0$ then produces the first-order correction and the second-order correction. The gradient-normalization step completes the pipeline: normalize the gradient of the approximate distance and solve a Poisson equation to project the result back onto the space of distance-like functions.","core_discovery":"The central claim is that for a bounded planar domain $\\Omega$, the distance to the boundary can be estimated more accurately than the heat method's plain log formula by differentiating the screened Poisson solution with respect to $\\lambda$. With $v$ solving $-\\Delta v + \\lambda^2 v = 0$ in $\\Omega$ and $v = 1$ on $\\partial\\Omega$, the heat method uses $\\mathrm{dist} \\approx -(1/\\lambda) \\log v$. The paper's first correction is $\\mathrm{dist} \\approx -v'_\\lambda/v$, where $v'_\\lambda$ solves the equation obtained by differentiating once, and the second correction adds the term $-\\frac{\\lambda}{2}[v''_{\\lambda\\lambda}/v - (v'_\\lambda/v)^2]$. These formulas arise from a Taylor expansion of $u = -(1/\\lambda)\\log v$ in $s = 1/\\lambda$ near $s = 0$, and numerically, after the standard gradient-normalization projection step, the second-order formula gives the smallest $L^2$ and $L^\\infty$ errors on all tested 2D shapes. The paper presents the formulas as practical improvements while noting that their rigorous asymptotic justification is not yet established.","pith_inferences":["The same $\\lambda$-differentiation step could be applied to any distance estimator expressible as a Laplace-type integral, not just the screened Poisson equation; testing this on other integral kernels would show how general the mechanism is.","The flat error curve for the second-order formula suggests the second-order term is removing a systematic bias rather than random noise; a bias-variance decomposition on a domain with known exact distance could confirm this and guide the choice of $\\lambda$.","A rigorous version of the Taylor extrapolation would need a bound on the remainder of the expansion; for convex domains with smooth boundaries, Green's function asymptotics might give an explicit $O(1/\\lambda^3)$ estimate that would settle when the second-order formula converges.","The reported tests are planar, so the practical payoff for surfaces and geodesics remains open; if the formulas extend, their parameter insensitivity would be a stronger advantage there than the small absolute accuracy gain seen in 2D."],"forward_implications":["On 2D shapes, both new differential formulas reduce $L^2$ and $L^\\infty$ distance errors relative to the heat method estimate; with gradient normalization, the second-order formula is the most accurate and the first-order formula is second.","The $L^2$ error of the second-order estimate stays nearly flat as $t = 1/\\lambda^2$ varies from 0.2 to 10, so the second-order estimate is far less sensitive to the user-chosen parameter than the plain heat method.","The Laplace-blended convolutional estimate, formed by weighting the LogConv and SoftMin formulas with the weights derived from Laplace asymptotics, improves accuracy over either convolutional formula taken alone on the tested shape.","Where the regularized solution is already the exact distance function, the three formulas coincide, so the observed improvements are concentrated in regions where the heat solution departs from the true distance, such as near skeleton branch points."],"supporting_citations":[{"why":"Supplies the LogConv and SoftMin convolutional distance transforms whose blend with Laplace weights is one of the two proposed improvements.","marker":"[9]"},{"why":"Provides the Laplace asymptotics that justify combining the two convolutional estimates with the proposed weights.","marker":"[10]"},{"why":"Establishes the heat-kernel asymptotics that underpin the heat method used as the baseline for the differential formulas.","marker":"[11]"},{"why":"Introduces the heat method and its gradient-normalization Poisson projection, which the new formulas are compared against and combined with.","marker":"[12]"},{"why":"Provides the later heat-method description whose time-step convention $\\lambda = 1/\\sqrt{t}$ the experiments adopt.","marker":"[13]"}],"fun_headline_variants":["Second-order Poisson trick cuts distance errors","Differentiate screened Poisson for sharper distances","Taylor fix beats heat method on 2D shapes","Higher-order Poisson formula refines distance maps","Screened Poisson derivative improves distance estimate"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The new differential formulas assume that the solution of the regularized equation is smooth enough in $1/\\lambda$ that its first- or second-order Taylor polynomial is accurate with no bound on the leftover error; the paper explicitly acknowledges that this assumption currently lacks mathematical justification.","fun_headline_variants_meta":{"raw":{"variants":["Second-order Poisson trick cuts distance errors","Differentiate screened Poisson for sharper distances","Taylor fix beats heat method on 2D shapes","Higher-order Poisson formula refines distance maps","Screened Poisson derivative improves distance estimate"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00017,"raw_usage":{"total_tokens":1216,"prompt_tokens":841,"completion_tokens":375,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":457,"completion_tokens_details":{"reasoning_tokens":319}},"tokens_in":457,"tokens_out":375,"duration_ms":3931,"temperature":1.0,"reasoning_tokens":319,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:12:58.548861+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Use a domain with an exactly known distance function, such as a disk or annulus, solve the screened Poisson equation and its $\\lambda$-derivative equations numerically for many values of $\\lambda$, and compare the three formulas with the exact distance; if the second-order formula does not consistently beat the first-order one as $\\lambda$ grows, or if the improvement disappears when the heuristic gradient-normalization step is removed, the central accuracy claim would be disproved.","supporting_citations":[{"cited_title":"Fast convolutional distance transform,","cited_arxiv_id":null,"evidence_quote":"Supplies the LogConv and SoftMin convolutional distance transforms whose blend with Laplace weights is one of the two proposed improvements."},{"cited_title":"On the behavior of the fundamental solution of the heat equation with variable coefficients,","cited_arxiv_id":null,"evidence_quote":"Establishes the heat-kernel asymptotics that underpin the heat method used as the baseline for the differential formulas."},{"cited_title":"Geodesics in heat: A new approach to computing distance based on heat flow,","cited_arxiv_id":null,"evidence_quote":"Introduces the heat method and its gradient-normalization Poisson projection, which the new formulas are compared against and combined with."},{"cited_title":"The heat method for distance computation,","cited_arxiv_id":null,"evidence_quote":"Provides the later heat-method description whose time-step convention $\\lambda = 1/\\sqrt{t}$ the experiments adopt."}],"review_version":1}