Pith. sign in

REVIEW 4 major objections 6 minor 67 references

Noise Matters: Diffusion Model-based Urban Mobility Generation with Collaborative Noise Priors

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read The paper claims that replacing i.i.d. noise with collaborative noise priors makes diffusion-based urban mobility generation match both individual trajectories and collective flows, with an average improvement of over 32% across six…

desk verdict Clever noise-prior construction, but the headline flow gains are partly baked in because the same flow matrix that builds the prior is also the evaluation target. read the letter →

arxiv 2412.05000 v2 pith:CPXVOSAS submitted 2024-12-06 cs.LG

classification cs.LG
keywords urbanmobilitygenerationdiffusionmodelscollaborativenoisepriorscollectiveflowssynthetictrajectoriesprivacy-preservingdatatrajectory
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

CoDiffMob is trying to show that the noise a diffusion model starts from is the decisive variable for generating realistic urban mobility, not just the denoiser architecture. The paper argues that standard i.i.d. Gaussian noise ignores the spatiotemporal correlations and social interactions in human movement, so it builds collaborative noise priors that fuse random noise with rule-based transition sequences steered by aggregate flow data. If the central claim is right, synthetic trajectories could replace privacy-sensitive, expensive web-derived mobility data in downstream tasks while preserving both individual-level statistics and city-wide flow patterns. The paper reports that this approach outperforms six baselines by over 32% on average and that the generated data resists membership-inference attacks.

What carries the argument

The central object is the collaborative noise prior, defined as $z_P = \mathrm{RhythmicBN}(z_{\mathrm{iid}} + z_F)$, where $z_F$ is obtained by mapping a flow-conditioned EPR transition sequence through inverse DDIM, and $z_{\mathrm{iid}}$ is standard Gaussian noise. The EPR (exploration and preferential return) model supplies individual-level actions — stay, return home, revisit, explore — and the exploration step is redirected by a flow-based policy drawn from the collective origin-destination matrix, so population patterns enter at the sampling level. Rhythmic batch normalization then rescales the per-time-step variance of the noise batch to be proportional to the crowd's moving probability, injecting the city's daily rhythm. The mechanism does the argument's work because DDIM sampling is deterministic: once the prior is fixed, the trajectory is fixed, so embedding collective knowledge into the starting noise is what lets a single denoiser reproduce both individual preferences and aggregate flows.

What would settle it

Build the flow matrix and population distribution for the noise prior from the first half of the observation week, evaluate CPC and MAPE on the second half, and compare with baselines trained the same way. If CoDiffMob's flow metrics drop to baseline levels, the claim that it learns collective mobility patterns would be refuted, since the apparent improvement would come from feeding the test-period statistics into the prior.

Watch

Extended reading notes

Core claim

The paper's central claim is that the initial noise fed into a DDIM trajectory generator is not a neutral random canvas: it already encodes movement semantics, and therefore the choice of noise is a primary modeling decision for urban mobility generation. Using inverse DDIM on real trajectories, the paper reports that noise-space direction tracks real-world moving direction with R²=0.9552 and that per-time-step noise variance follows the crowd's daily movement probability. On this basis CoDiffMob constructs a collaborative noise prior by sampling location-transition sequences with an EPR-style rule-based model that is steered by the collective flow matrix and population distribution, mapping those sequences into noise space through inverse DDIM, adding white noise, and applying rhythmic batch normalization so that noise variance matches the movement rhythm. Starting denoising from this prior, the model is claimed to generate trajectories that match individual statistics and collective origin-destination flows simultaneously, with average improvement over 32% versus six baselines and with privacy preserved.

Load-bearing premise

The load-bearing premise is that the collective flow matrix and population distribution used to build the noise prior come from the same real trajectories that are later used to compute the flow evaluation metrics; the paper never shows the prior is built from held-out data, so the reported flow improvements may partly reflect the evaluation statistics being fed into the generator.

Editorial extensions

If this is right

  • Synthetic trajectories generated this way can substitute for real trajectories in training mobility predictors; the paper reports that mixing CoDiffMob output with real data improves accuracy relative to mixing DiffTraj output or using real data alone.
  • The reported flow metrics (CPC above 0.6 and MAPE below 0.7 on both datasets) mean a single generative model can reproduce aggregate origin-destination structure, not just individual-level statistics.
  • Privacy checks indicate the model produces novel trajectories rather than memorized ones: over 90% of generated trajectories have no real counterpart with more than 40% overlap, and membership-inference success stays below 0.6.
  • The ablation results show that removing the noise prior degrades flow fidelity sharply (CPC drops from 0.68 to 0.34 on the ISP dataset), so the prior is the component doing the collective-pattern work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper computes the collective flow matrix from the same real dataset that supplies the evaluation flows; a cleaner test would build the prior from a separate time window or region and evaluate on held-out flows, to isolate how much of the reported improvement is prior injection versus learned generation.
  • The observed correlation between movement direction and noise-space direction suggests noise-prior editing could become a general steering mechanism for spatiotemporal sequence generation, letting practitioners impose direction or timing constraints by reshaping the starting noise without retraining.
  • A natural stress test is to feed CoDiffMob a deliberately wrong or corrupted flow matrix and measure whether generated flows move toward that wrong matrix; if they do, the prior dominates generation, which is both a design feature and a risk for misleading synthetic data.
  • The rhythmic batch-normalization trick points to a cheap way to inject external temporal signals such as weather or special events into any trajectory diffusion model by rescaling per-step noise variance instead of changing the network.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes CoDiffMob, a diffusion model for urban mobility trajectory generation that constructs a 'collaborative noise prior' by combining an EPR-based rule-based sampler with the dataset's collective flow matrix F, fusing the resulting informative noise with i.i.d. Gaussian noise, and applying a rhythmic batch normalization based on the dataset's hourly movement probability. The authors claim that this prior improves both individual trajectory realism and collective flow reproduction, reporting an average improvement of over 32% over six baselines, plus privacy and downstream-utility experiments. The paper includes motivation based on observed correlations between initial noise and generated trajectories, and it releases code and data.

Significance. The idea of injecting collective mobility statistics into the noise space of a diffusion model is timely and potentially useful for synthetic mobility generation, especially for urban computing and privacy-preserving data sharing. The manuscript's strengths include a clear problem formulation, a concrete algorithmic specification (Algorithm 1), public code and data, and an explicit privacy evaluation. However, the central quantitative claim—the >32% improvement, driven mainly by flow metrics—is currently undermined by evaluation leakage: the same flow matrix F used to construct the noise prior is also the target of the CPC/MAPE metrics. The ablation study further shows that the rule-based flow sampler alone achieves better flow metrics than the full diffusion model, so the reported flow improvements do not demonstrate that the diffusion model learned collective dynamics. If the evaluation is repaired with held-out flow statistics, proper baselines, and uncertainty quantification, the contribution could be solid, but as presented the headline claim is not established.

major comments (4)
  1. [§4.2, Algorithm 1, and §5.2, Eqs. (15)-(16)] The evaluation of collective-flow quality is circular. The collaborative noise prior is built from the real dataset's flow matrix F and population P (Eq. 7, Algorithm 1), and the CPC/MAPE metrics measure agreement between the generated aggregate flow and that same F (Eqs. 15-16). The reported >35% flow improvement over baselines therefore partly reflects direct injection of the target statistics into the sampling process rather than the diffusion model's learned representation. Please re-run the flow evaluation with a held-out temporal or spatial split to compute F, and give baselines access to the same flow statistics, or compare against a baseline that also samples transitions directly from F.
  2. [§5.5, Table 2] The ablation results undermine the claim that the diffusion model captures collective patterns. On ISP, the 'w/o Noise Fusion' variant—which relies only on the collaborative transition sampling process—achieves CPC=0.7165 and MAPE=0.5351, which are better than full CoDiffMob's CPC=0.6842 and MAPE=0.6361. Thus the rule-based flow prior alone reproduces the flow matrix more accurately than the full model, and the improvement over baselines is attributable to direct sampling from F rather than to the learned denoiser. Please report flow metrics for the pure rule-based sampler and for the diffusion model alone, and discuss the resulting trade-off explicitly.
  3. [§5.2, Table 1] Individual trajectory metrics (Radius, Distance, Duration, DailyLoc) are reported as single Kolmogorov-Smirnov statistics without confidence intervals, sample sizes, or statistical significance tests. Several apparent improvements over DiffTraj are small (e.g., ISP Distance 0.1203 vs. 0.1662; MME Duration 0.0796 vs. 0.0949). Without uncertainty quantification across multiple generation seeds, the claim of consistent superiority on individual-level distributions is not supported. Please provide error bars and significance tests, such as paired bootstrap or repeated KS tests.
  4. [§5.4, Figure 6] The downstream utility experiment shows accuracy values on a 0-0.3 scale, and the differences between Real+DiffTraj and Real+CoDiffMob are small. No error bars or significance tests are reported. The conclusion that CoDiffMob 'leads to more significant improvements in downstream task performance' is therefore not supported by the evidence as presented. Please report mean and standard deviation over repeated runs and a statistical test of the difference.
minor comments (6)
  1. [Abstract] The sentence 'In this paper, we propose CoDiffMob, a diffusion model for urban mobility generation with collaborative noise priors, we emphasize the critical role of noise' is a comma splice; consider '...collaborative noise priors. We emphasize...'.
  2. [§4.2.2, Eq. (8)] The notation for the noise prior is inconsistent: z_p and z_P are both used, and 'z_i.i.d.' is not typeset consistently. Please unify the subscript and define all symbols.
  3. [Algorithm 1 and Eq. (7)] Algorithm 1 calls π with additional arguments (n_ω, β1, β2, P, ρ, γ, F) that do not appear in the definition of π in Eq. (7). Please align the notation.
  4. [Figure 3 and surrounding text] The caption labels panel (c) as 'DiffTraj' and reports CPC=0.3992 for both (a) and (c); the text says 'the image on the right shows the result generated by the best baseline, Difftraj.' This appears to be a typo in the CPC value or the caption.
  5. [Appendix B.2, Eq. (15)] In the definition of N_CC, 'min(FX,FY)' should be the element-wise minimum min(FX_ij, FY_ij) over matrix entries.
  6. [§5.2] The statements 'average improvement of over 40%' and 'over 31%' for trajectory metrics are not derivable from Table 1 as presented; please specify the exact aggregation formula and the baseline used for each percentage.

Circularity Check

1 steps flagged · score 6.0 of 10

Collective-flow claim is partially circular: the noise prior is constructed from the same real flow matrix F that later defines the CPC/MAPE evaluation target.

  1. fitted input called prediction [Section 4.2.1 (Eq. 7), Algorithm 1, and Section 5.1.2/Eqs. 15-16]
    "When an individual chooses to explore new locations, the strategy π follows the flow-based policy π_F(l_i | \tilde{F}_{l_{i-1}}) and select target location based on the distribution of collective movements \tilde{F}_{l_{i-1}}, which is computed from flow F_{l_{i-1}}. Algorithm 1 then sets l_i = π(...) with collective flows F, computes z_F = Inverse-DDIM(x_F), and uses z_p = Rhythmic BN(z_i.i.d. + z_F) as initial noise. Evaluation: Given real-world dataset X and generated data Y, we calculate collective flows F_X and F_Y from them, CPC is calculated by Eq. (15) and MAPE by Eq. (16)."

    The collaborative transition policy (Eq. 7) samples locations from the real dataset's own flow matrix F, and Algorithm 1 maps those sampled sequences into the diffusion noise prior via Inverse-DDIM. The CPC/MAPE metrics (Eqs. 15-16) then measure the generated flow F_Y against that same real matrix F_X. High CPC / low MAPE therefore reflects injection of the target statistics into the prior, not an independent discovery of collective mobility by the diffusion model. Baselines receive no such target-flow injection, so the comparison is not controlled. The ablation in Table 2 supports this: removing the diffusion component ('w/o Noise Fusion') leaves the F-based rule and yields even higher CPC (0.7165 vs 0.6842), showing flow fidelity is carried by the input F, not by learned dynamics.

full rationale

The paper is not wholly circular: Radius, Distance, Duration, DailyLoc, diversity, privacy, and downstream-utility results are evaluated against trajectory statistics rather than against the flow matrix F injected into the prior, so those parts retain independent content. However, the abstract's central claim of 'capturing collective patterns' with 'over 32% improvement' rests substantially on CPC/MAPE, and those collective-flow metrics are computed against the same F_X that the collaborative noise prior consumes (Eq. 7, Algorithm 1). The paper does not state that F is computed from held-out data, and the ablation shows the F-based rule alone outperforms the full model on CPC, confirming that the flow gain is inherited from the input statistics. This is a fitted-input-called-prediction pattern: the evaluation target is built into the generative prior. No load-bearing self-citation or uniqueness-theorem circularity was found. Score 6 reflects partial, rather than total, circularity: the collective-flow 'prediction' reduces by construction, while individual-level claims still stand on independent evidence.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claim rests on data-derived inputs (flow matrix F, population P, hourly movement probability P_t_move) that are injected into the noise prior and also serve as evaluation targets. The EPR rule-based sampler is parameterized by undisclosed fitted values, and several components (noise fusion, rhythmic batch norm, collaborative mixing) are combined without isolated ablations for each design choice.

free parameters (3)
  • EPR model parameters = not reported
    Algorithm 1 requires {n_omega, beta1, beta2, P, rho, gamma}; these control stay/return/explore probabilities. Their values are never given; if tuned per dataset, they inject fitted information into the noise prior.
  • Rhythmic BatchNorm scaling R_t = proportional to P_t_move from the real dataset
    Equation 9 sets the noise variance at each time to match the dataset's hourly moving probability. This is a data-derived per-time scaling, effectively a fitted temporal profile.
  • collaborative mixing in Eq. 7 = not specified
    The combination of individual policy pi_I and flow policy pi_F is not formally defined; the precedence or weight is a free design choice that affects generated trajectories.
assumptions (6)
  • standard math Gaussian forward process and score-based denoising assumptions of DDPM/DDIM (Eq. 1-3)
    The method builds on standard diffusion theory; these are accepted background results.
  • domain assumption DDIM inversion of a rule-based trajectory yields a noise vector that carries the trajectory's semantic content
    Section 4.2.2 assumes Inverse-DDIM(z_F) preserves the collective transition information from x_F after fusion with white noise. The paper provides empirical motivation but no guarantee.
  • domain assumption The EPR rule-based model augmented with flow policy generates plausible individual trajectories
    Section 4.2.1 relies on EPR to produce x_F; the fidelity of the noise prior depends on this simulation model.
  • domain assumption Aggregate flow matrix F and population P are available and accurate at inference time
    Algorithm 1 requires these inputs; their source and representativeness are not discussed, and the same F is used as evaluation target.
  • domain assumption Noise variance proportional to P_t_move improves temporal realism
    Section 4.2.3 introduces Rhythmic BatchNorm without an ablation isolating this component.
  • domain assumption The two privacy tests (uniqueness overlap and membership inference) establish privacy protection
    Section 5.3 treats attack success below 0.6 and overlap below 40% as sufficient, without comparison to baselines or formal privacy guarantees.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Noise Matters: Diffusion Model-based Urban Mobility Generation with Collaborative Noise Priors." pith.science (2026). https://pith.science/paper/CPXVOSAS

@misc{pith2026241205000,
  author       = {Pith},
  title        = {Pith review of: Noise Matters: Diffusion Model-based Urban Mobility Generation with Collaborative Noise Priors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CPXVOSAS}},
  note         = {Machine review of arXiv:2412.05000}
}
read the original abstract

With global urbanization, the focus on sustainable cities has largely grown, driving research into equity, resilience, and urban planning, which often relies on mobility data. The rise of web-based apps and mobile devices has provided valuable user data for mobility-related research. However, real-world mobility data is costly and raises privacy concerns. To protect privacy while retaining key features of real-world movement, the demand for synthetic data has steadily increased. Recent advances in diffusion models have shown great potential for mobility trajectory generation due to their ability to model randomness and uncertainty. However, existing approaches often directly apply identically distributed (i.i.d.) noise sampling from image generation techniques, which fail to account for the spatiotemporal correlations and social interactions that shape urban mobility patterns. In this paper, we propose CoDiffMob, a diffusion model for urban mobility generation with collaborative noise priors, we emphasize the critical role of noise in diffusion models for generating mobility data. By leveraging both individual movement characteristics and population-wide dynamics, we construct novel collaborative noise priors that provide richer and more informative guidance throughout the generation process. Extensive experiments demonstrate the superiority of our method, with generated data accurately capturing both individual preferences and collective patterns, achieving an improvement of over 32%. Furthermore, it can effectively replace web-derived mobility data to better support downstream applications, while safeguarding user privacy and fostering a more secure and ethical web. This highlights its tremendous potential for applications in sustainable city-related research. The code and data are available at https://github.com/tsinghua-fib-lab/CoDiffMob.

Figures

Figures reproduced from arXiv: 2412.05000 by the authors.

Figure 1
Figure 1. Correlations between noise and trajectories. (a) shows the t-SNE plot of the noise corresponding to the trajectories. (b) [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The overall workflow of mobility generation with collaborative noise priors. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Visualization of the observed flows and generated flows on ISP dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Success ratio of membership inference attack using [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Mobility prediction based on generated trajectories. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Visualization of the trajectory distribution of ISP dataset at different timestamps. [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Visualization of the real and generated flow on MME dataset. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 33 canonical work pages

  1. [1]

    Vincent Bindschaedler and Reza Shokri. 2016. Synthesizing plausible privacy- preserving location traces. In 2016 IEEE Symposium on Security and Privacy (SP). IEEE, 546–563

  2. [2]

    Chen Chu, Hengcai Zhang, Peixiao Wang, and Feng Lu. 2024. Simulating hu- man mobility with a trajectory generation framework based on diffusion model. International Journal of Geographical Information Science 38, 5 (2024), 847–878

  3. [3]

    Florinel-Alin Croitoru, Vlad Hondru, Radu Tudor Ionescu, and Mubarak Shah

  4. [4]

    Prafulla Dhariwal and Alexander Nichol. 2021. Diffusion models beat gans on image synthesis. Advances in neural information processing systems 34 (2021), 8780–8794

  5. [5]

    Umberto Fattore, Marco Liebsch, Bouziane Brik, and Adlen Ksentini. 2020. AutoMEC: LSTM-based user mobility prediction for service management in distributed MEC resources. In Proceedings of the 23rd International ACM Conference on Modeling, Analysis and Simulation of Wireless and Mobile Systems. 155–159

  6. [6]

    Huifang Feng, Fengshan Bai, and Youji Xu. 2019. Identification of critical roads in urban transportation network based on GPS trajectory data.Physica A: Statistical Mechanics and its Applications 535 (2019), 122337

  7. [7]

    Jie Feng, Zeyu Yang, Fengli Xu, Haisu Yu, Mudan Wang, and Yong Li. 2020. Learning to simulate human mobility. In Proceedings of the 26th ACM SIGKDD international conference on knowledge discovery & data mining. 3426–3433

  8. [8]

    Songwei Ge, Seungjun Nah, Guilin Liu, Tyler Poon, Andrew Tao, Bryan Catanzaro, David Jacobs, Jia-Bin Huang, Ming-Yu Liu, and Yogesh Balaji. 2023. Preserve your own correlation: A noise prior for video diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 22930–22941

Show all 67 references
  1. [9]

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2020. Generative adversarial networks. Commun. ACM 63, 11 (2020), 139–144

  2. [10]

    Andrea Hess, Karin Anna Hummel, Wilfried N Gansterer, and Günter Haring

  3. [11]

    Jonathan Ho and Stefano Ermon. 2016. Generative adversarial imitation learning. Advances in neural information processing systems 29 (2016)

  4. [12]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems 33 (2020), 6840– 6851

  5. [13]

    Sibren Isaacman, Richard Becker, Ramón Cáceres, Margaret Martonosi, James Rowland, Alexander Varshavsky, and Walter Willinger. 2012. Human mobil- ity modeling at metropolitan scales. In Proceedings of the 10th international conference on Mobile systems, applications, and servi...

  6. [14]

    Shan Jiang, Yingxiang Yang, Siddharth Gupta, Daniele Veneziano, Shounak Athavale, and Marta C González. 2016. The TimeGeo modeling framework for urban mobility without travel surveys. Proceedings of the National Academy of Sciences 113, 37 (2016), E5370–E5378

  7. [15]

    Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. 2022. Elucidating the Design Space of Diffusion-Based Generative Models. arXiv:2206.00364 [cs.CV] https://arxiv.org/abs/2206.00364

  8. [16]

    Diederik P Kingma. 2013. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114 (2013)

  9. [17]

    Yong Li, Yuan Yuan, Jingtao Ding, and Depeng Jin. 2023. Learning the complexity of urban mobility with deep generative collaboration network. (2023)

  10. [18]

    Zinan Lin, Alankar Jain, Chen Wang, Giulia Fanti, and Vyas Sekar. 2020. Using GANs for Sharing Networked Time Series Data: Challenges, Initial Promise, and Open Questions. In Proceedings of the ACM Internet Measurement Conference (IMC ’20). ACM, 464–483. https://doi.org/10.114...

  11. [19]

    Kang Liu, Xin Jin, Shifen Cheng, Song Gao, Ling Yin, and Feng Lu. 2024. Act2Loc: a synthetic trajectory generation method by combining machine learning and mechanistic models. International Journal of Geographical Information Science 38, 3 (2024), 407–431

  12. [20]

    Qingyue Long, Huandong Wang, Tong Li, Lisi Huang, Kun Wang, Qiong Wu, Guangyu Li, Yanping Liang, Li Yu, and Yong Li. 2023. Practical synthetic human trajectories generation based on variational point processes. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Disc...

  13. [21]

    Qingyue Long, Yuan Yuan, and Yong Li. 2024. A Universal Model for Human Mobility Prediction. arXiv preprint arXiv:2412.15294 (2024)

  14. [22]

    Massimiliano Luca, Gianni Barlacchi, Bruno Lepri, and Luca Pappalardo. 2021. A survey on deep learning for human mobility. ACM Computing Surveys (CSUR) 55, 1 (2021), 1–44

  15. [23]

    Zhengxiong Luo, Dayou Chen, Yingya Zhang, Yan Huang, Liang Wang, Yu- jun Shen, Deli Zhao, Jingren Zhou, and Tieniu Tan. 2023. Videofusion: De- composed diffusion models for high-quality video generation. arXiv preprint arXiv:2303.08320 (2023)

  16. [24]

    Frank J Massey Jr. 1951. The Kolmogorov-Smirnov test for goodness of fit.Journal of the American statistical Association 46, 253 (1951), 68–78

  17. [25]

    2024.The Sustainable Development Goals Report 2024 (2024 ed.)

    United Nations Department of Economic and Social Affairs. 2024.The Sustainable Development Goals Report 2024 (2024 ed.). United Nations. https://www.un- ilibrary.org/content/books/9789213589755

  18. [27]

    Haonan Qiu, Menghan Xia, Yong Zhang, Yingqing He, Xintao Wang, Ying Shan, and Ziwei Liu. 2024. FreeNoise: Tuning-Free Longer Video Diffusion via Noise Rescheduling. arXiv:2310.15169 [cs.CV] https://arxiv.org/abs/2310.15169

  19. [28]

    Jinmeng Rao, Song Gao, Yuhao Kang, and Qunying Huang. 2020. LSTM-TrajGAN: A deep learning approach to trajectory privacy protection. arXiv preprint arXiv:2006.10521 (2020)

  20. [29]

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 10684–10695

  21. [30]

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceedings,...

  22. [31]

    Chenyang Shao, Fengli Xu, Bingbing Fan, Jingtao Ding, Yuan Yuan, Meng Wang, and Yong Li. 2024. Beyond imitation: Generating human mobility from context- aware reasoning with large language models. arXiv preprint arXiv:2402.09836 (2024)

  23. [32]

    Zhi Sheng, Yuan Yuan, Jingtao Ding, and Yong Li. 2025. Unveiling the Power of Noise Priors: Enhancing Diffusion Models for Mobile Traffic Prediction. arXiv preprint arXiv:2501.13794 (2025)

  24. [33]

    Reza Shokri, Marco Stronati, Congzheng Song, and Vitaly Shmatikov. 2017. Membership Inference Attacks Against Machine Learning Models. In 2017 IEEE Symposium on Security and Privacy (SP). 3–18. https://doi.org/10.1109/SP.2017. 41

  25. [34]

    Filippo Simini, Gianni Barlacchi, Massimilano Luca, and Luca Pappalardo. 2021. A deep gravity model for mobility flows generation. Nature communications 12, 1 (2021), 6576

  26. [35]

    Chaoming Song, Tal Koren, Pu Wang, and Albert-László Barabási. 2010. Modelling the scaling properties of human mobility. Nature physics 6, 10 (2010), 818–823

  27. [36]

    Jiaming Song, Chenlin Meng, and Stefano Ermon. 2020. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502 (2020)

  28. [37]

    Xuan Song, Hiroshi Kanasugi, and Ryosuke Shibasaki. 2016. Deeptransport: Prediction and simulation of human mobility and transportation mode at a citywide level. In Proceedings of the twenty-fifth international joint conference on artificial intelligence. 2618–2624

  29. [39]

    Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole

    Yang Song, Jascha Sohl-Dickstein, Diederik P. Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. 2021. Score-Based Generative Modeling through Stochas- tic Differential Equations. arXiv:2011.13456 [cs.LG] https://arxiv.org/abs/2011. 13456

  30. [40]

    Hongyuan Su, Yu Zheng, Jingtao Ding, Depeng Jin, and Yong Li. 2024. Met- roGNN: Metro Network Expansion with Reinforcement Learning. In Companion Proceedings of the ACM on Web Conference 2024. 650–653

  31. [41]

    Zhen Tu, Kai Zhao, Fengli Xu, Yong Li, Li Su, and Depeng Jin. 2019. Protecting Trajectory From Semantic Attack Considering𝑘 -Anonymity,𝑙 -Diversity, and𝑡 -Closeness. IEEE Transactions on Network and Service Management 16, 1 (2019), 264–278. https://doi.org/10.1109/TNSM.2018.2877790

  32. [42]

    Narek Tumanyan, Michal Geyer, Shai Bagon, and Tali Dekel. 2022. Plug- and-Play Diffusion Features for Text-Driven Image-to-Image Translation. arXiv:2211.12572 [cs.CV] https://arxiv.org/abs/2211.12572

  33. [43]

    Narek Tumanyan, Michal Geyer, Shai Bagon, and Tali Dekel. 2023. Plug-and-play diffusion features for text-driven image-to-image translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 1921– 1930

  34. [44]

    Laurens Van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-SNE. Journal of machine learning research 9, 11 (2008)

  35. [45]

    A Vaswani. 2017. Attention is all you need. Advances in Neural Information Processing Systems (2017)

  36. [46]

    Huandong Wang, Changzheng Gao, Yuchen Wu, Depeng Jin, Lina Yao, and Yong Li. 2023. PateGail: a privacy-preserving mobility trajectory generator with imitation learning. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 37. 14539–14547

  37. [47]

    Fengli Xu, Zhen Tu, Yong Li, Pengyu Zhang, Xiaoming Fu, and Depeng Jin

  38. [48]

    Ruihan Yang, Prakhar Srivastava, and Stephan Mandt. 2023. Diffusion probabilis- tic modeling for video generation. Entropy 25, 10 (2023), 1469. WWW ’25, April 28-May 2, 2025, Sydney, NSW, Australia Yuheng Zhang, Yuan Yuan, Jingtao Ding, Jian Yuan, and Yong Li

  39. [49]

    Zhongchao Yi, Zhengyang Zhou, Qihe Huang, Yanjiang Chen, Liheng Yu, Xu Wang, and Yang Wang. 2024. Get Rid of Isolation: A Continuous Multi-task Spatio-Temporal Learning Framework. InThe Thirty-eighth Annual Conference on Neural Information Processing Systems

  40. [50]

    Yuan Yuan, Jingtao Ding, Jie Feng, Depeng Jin, and Yong Li. 2024. Unist: A prompt-empowered universal model for urban spatio-temporal prediction. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 4095–4106

  41. [51]

    Yuan Yuan, Jingtao Ding, Chonghua Han, Depeng Jin, and Yong Li. 2024. A Foundation Model for Unified Urban Spatio-Temporal Flow Prediction. arXiv preprint arXiv:2411.12972 (2024)

  42. [52]

    Yuan Yuan, Jingtao Ding, Chenyang Shao, Depeng Jin, and Yong Li. 2023. Spatio- temporal diffusion point processes. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 3173–3184

  43. [53]

    Yuan Yuan, Jingtao Ding, Huandong Wang, and Depeng Jin. 2024. Generating Daily Activities with Need Dynamics. ACM Transactions on Intelligent Systems and Technology 15, 2 (2024), 1–28

  44. [54]

    Yuan Yuan, Jingtao Ding, Huandong Wang, Depeng Jin, and Yong Li. 2022. Activ- ity trajectory generation via modeling spatiotemporal dynamics. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 4752–4762

  45. [55]

    Yuan Yuan, Chonghua Han, Jingtao Ding, Depeng Jin, and Yong Li. 2024. Urbandit: A foundation model for open-world urban spatio-temporal learning. arXiv preprint arXiv:2411.12164 (2024)

  46. [56]

    Yuan Yuan, Huandong Wang, Jingtao Ding, Depeng Jin, and Yong Li. 2023. Learning to simulate daily activities via modeling dynamic human needs. In Proceedings of the ACM Web Conference 2023. 906–916

  47. [57]

    Jing Zhang, Qihan Huang, Yirui Huang, Qian Ding, and Pei-Wei Tsai. 2023. DP- TrajGAN: A privacy-aware trajectory generation model with differential privacy. Future Generation Computer Systems 142 (2023), 25–40

  48. [58]

    Yunke Zhang, Yuming Lin, Guanjie Zheng, Yu Liu, Nicholas Sukiennik, Fengli Xu, Yongjun Xu, Feng Lu, Qi Wang, Yuan Lai, et al. 2025. MetaCity: Data-driven sustainable development of complex cities. The Innovation (2025)

  49. [59]

    Zhongwei Zhang, Fuchen Long, Yingwei Pan, Zhaofan Qiu, Ting Yao, Yang Cao, and Tao Mei. 2024. TRIP: Temporal Residual Learning with Image Noise Prior for Image-to-Video Diffusion Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 8671–8681

  50. [60]

    Yu Zheng, Licia Capra, Ouri Wolfson, and Hai Yang. 2014. Urban computing: concepts, methodologies, and applications. ACM Transactions on Intelligent Systems and Technology (TIST) 5, 3 (2014), 1–55

  51. [61]

    Yu Zheng, Qianyue Hao, Jingwei Wang, Changzheng Gao, Jinwei Chen, Depeng Jin, and Yong Li. 2024. A Survey of Machine Learning for Urban Decision Making: Applications in Planning, Transportation, and Healthcare. Comput. Surveys (2024)

  52. [62]

    Yu Zheng, Yuming Lin, Liang Zhao, Tinghai Wu, Depeng Jin, and Yong Li. 2023. Spatial planning of urban communities via deep reinforcement learning. Nature Computational Science 3, 9 (2023), 748–762

  53. [63]

    Yu Zheng, Hongyuan Su, Jingtao Ding, Depeng Jin, and Yong Li. 2023. Road planning for slums via deep reinforcement learning. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. 5695– 5706

  54. [64]

    Yu Zheng, Lizhu Zhang, Xing Xie, and Wei-Ying Ma. 2009. Mining interesting locations and travel sequences from GPS trajectories. In Proceedings of the 18th international conference on World wide web. 791–800

  55. [65]

    Zhengyang Zhou, Qihe Huang, Binwu Wang, Jianpeng Hou, Kuo Yang, Yuxuan Liang, and Yang Wang. 2024. ComS2T: A complementary spatiotemporal learn- ing system for data-adaptive model evolution. arXiv preprint arXiv:2403.01738 (2024)

  56. [66]

    Yuanshao Zhu, Yongchao Ye, Shiyao Zhang, Xiangyu Zhao, and James Yu. 2023. Difftraj: Generating gps trajectory with diffusion probabilistic model. Advances in Neural Information Processing Systems 36 (2023), 65168–65188. A Methodology Details A.1 Diffusion Denoiser Following [...

  57. [2015]

    ACM Computing Surveys (CSUR) 48, 3 (2015), 1–39

    Data-driven human mobility modeling: a survey and engineering guidance for mobile networking. ACM Computing Surveys (CSUR) 48, 3 (2015), 1–39

  58. [2017]

    InProceedings of the 26th international conference on world wide web

    Trajectory recovery from ash: User privacy is not preserved in aggregated mobility data. InProceedings of the 26th international conference on world wide web. 1241–1250

  59. [2023]

    IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 9 (2023), 10850–10869

    Diffusion models in vision: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence 45, 9 (2023), 10850–10869

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.