Pith. sign in

REVIEW 2 major objections 6 minor 2 cited by

Network-Wide Traffic Flow Estimation Across Multiple Cities with Global Open Multi-Source Data: A Large-Scale Case Study in Europe and North America

T0 review · 2 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

Pith's one-line read Using only globally available map images and observed speeds, an attention-based graph neural network estimates network-wide traffic flow in 15 European and North American cities with average SMAPE of 22.81%, beating standard baselines in…

desk verdict A useful 15-city benchmark and a sensible map-fusion idea, but the central claim about estimating flow on unmonitored roads is not tested because the sensor-distribution map marks every target sensor. read the letter →

arxiv 2502.03798 v1 pith:URFSZNMH submitted 2025-02-06 cs.LG

classification cs.LG
keywords network-widetrafficflowcross-cityestimationglobalopenmulti-sourcedatagraphneuralnetworkmapimagesattentionmechanismsensorcoverage
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

The paper claims that the usual trade-off between accuracy and generality in network-wide traffic flow estimation can be broken by feeding a graph neural network three publicly available map images: OpenStreetMap, a sensor-location map, and a population density map, together with observed speed data. Across 15 cities in Europe and North America, the model estimates traffic flow on roads not used in training with an average symmetric mean absolute percentage error of 22.81%, and it beats four baselines in every city. The authors argue that map images capture the causes and consequences of traffic—land use, road topology, points of interest, and human population—so they carry information that speed alone lacks. If correct, cities with sparse sensor coverage could obtain network-wide flow estimates from open data, and one model could transfer across cities without city-specific data collection.

What carries the argument

The core machinery is a pair of map images (OpenStreetMap, sensor distribution map, population density map) encoded with residual convolution blocks, fused by a Triple Cross-Attention Block (TCAB) that computes cross-attention maps between each pair of image types, and further condensed by a Dense Connection Block (DCB) that repeatedly concatenates and re-encodes features. On the graph side, speed observations are passed through a graph encoder, a spatial attention module (GRU then GAT), and a temporal attention module (GAT then GRU), producing node embeddings. The image embedding and node embedding are concatenated and decoded by an MLP to yield flow per road per time step. The key idea is that the three map types encode static geographical and demographical context that is either a cause or consequence of traffic, while speed captures dynamics, and the cross-attention fuses the two domains.

What would settle it

Train on cities where the sensor distribution map omits the target road's own marker, or test on genuinely unmonitored roads, and compare SMAPE. If accuracy degrades markedly when the target's red dot is absent, the learned mapping depends on the marker rather than on surrounding context, and the method would not transfer to truly unmonitored networks.

Watch

Extended reading notes

Core claim

The central discovery is that GOMS map images, used as a unified data format, allow a single attention-based graph neural network to estimate traffic flow on unobserved road segments with stable accuracy across multiple cities. The paper constructs the sensor distribution map by drawing known sensor locations as red dots on a background map, and combines it with OpenStreetMap images and satellite-derived population density maps. A Triple Cross-Attention Block computes pairwise attention between the three encoded images, and a Dense Connection Block concatenates and re-processes the fused features; graph spatial and temporal attention modules encode speed observations. In experiments over 3,126 road segments in 15 cities (UTD19 for Europe, PeMS for California), the method yields an average SMAPE of 22.81%, with city-level SMAPEs between 16.79% and 26.93%, and outperforms spatial averaging, LSTM, GCMC, and GATv2 baselines in every city. The authors interpret the small spread between worst and best city errors as evidence that accuracy and generality need not trade off.

Load-bearing premise

The model assumes that a road whose location is not drawn on the sensor distribution map will still be estimated accurately, even though every road seen during training has its own red dot on that map.

Editorial extensions

If this is right

  • Cities with only open map data and a modest set of speed observations could produce network-wide flow estimates without installing dense sensor networks.
  • Because the input images are obtained from globally available sources, the same pipeline can be applied to new cities, enabling cross-city comparisons of traffic patterns.
  • The ablation study shows that combining all three map types gives the best and most stable errors, so adding further GOMS sources such as weather may improve accuracy further.
  • The stability of errors across cities with very different sizes and sensor densities supports the paper's claim that the accuracy–generality trade-off is not inherent.

Reading between the lines

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

  • The sensor distribution map may leak the identity of the target road, because the target itself is a sensor with a red dot; the paper never tests the truly unmonitored case, so the claimed generality may be partly an artifact of this input design.
  • A natural test is to erase the target's own marker from the sensor map during inference; a model that truly uses surrounding context should survive, while one that memorizes locations would fail.
  • The method still requires some observed speed data for the target city, so it is not purely map-based; the 'open data' claim is about the static inputs rather than the traffic observations.
  • If the map-based approach transfers, it could be extended to other urban phenomena that are caused or reflected by built environment and population, such as noise or air pollution estimation.
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

2 major / 6 minor

Summary. The manuscript proposes a network-wide traffic flow estimation (NTFE) method that fuses three map-image inputs (OpenStreetMap, a sensor distribution map, and a population density map) with observed speed data in an attention-based graph neural network. The method is evaluated on 15 cities in Europe and North America by training on a subset of sensor-equipped road segments and testing on held-out segments. The authors report stable performance, with an average SMAPE of 22.81%, and ablation experiments indicating that adding GOMS maps improves accuracy.

Significance. The scale of the study is a genuine strength: 15 cities, 3,126 road segments, a consistent evaluation protocol, and systematic ablations over data sources and network components. If the reported accuracy held for truly unmonitored roads, the GOMS-map representation would be a valuable step toward more transferable traffic-flow estimation. However, the experiments as designed do not establish that, because the sensor distribution map always contains a marker at the target node and all test targets are themselves sensor locations. The paper's central deployment scenario is therefore not evaluated, and the stated claim of breaking an accuracy-generality trade-off is stronger than the evidence supports.

major comments (2)
  1. [Section 5.1.2, Section 5.4] There is a train/serve mismatch for the sensor distribution map. The map is built by drawing red dots at sensor locations from the UTD19/PeMS coordinates (Section 5.1.2), and every graph node is a fixed sensor location (Section 3.2). Consequently, every training and testing image contains a red dot at the target node itself. For a truly unmonitored road, the target would have no such dot, and only neighboring sensors would appear. The model is therefore trained and evaluated on the conditional distribution P(map | target is a sensor), which is never realized for the deployment scenario the paper motivates. The ablations in Section 5.5.1 do not control for this, because every variant that includes the sensor map also includes the target dot. The reported average SMAPE of 22.81% therefore does not support the abstract's claim of estimating flow on unobserved, unmonitored roads. I ask the authors to report results with the target dot masked in the sensor distribution map at evaluation time, and ideally to evaluate on road segments that genuinely have no sensor; if the intended scope is only held-out sensor locations, the text should say so explicitly and temper the deployment claims.
  2. [Section 6, Section 5.4] The conclusion states that 'we currently train a separate neural network for each city to guarantee the estimation accuracy' (Section 6). The experiments therefore show that the same architecture and GOMS inputs can be calibrated per city with stable accuracy, but they do not show that a single model transfers across cities. The abstract's 'across cities' phrasing and the first contribution's reference to 'universal NTFE methods' are stronger than this evidence. If the intended claim is only that GOMS data are globally available, then the accuracy-generality trade-off is only partially addressed, because local observed flow data and per-city training are still required. I recommend either adding a cross-city transfer experiment (train on several cities, test on a held-out city) or revising the claims to describe globally available inputs with per-city calibration.
minor comments (6)
  1. [Abstract] The phrase 'we first time advocate' should be rewritten, for example as 'we advocate for the first time'.
  2. [Table 2] The GATv2 SMAPE for Bolton is reported as '3303%'; this appears to be a typo and should likely be '33.03%'.
  3. [Section 5.4] The sentence 'While the the X-axis represents...' contains a duplicated 'the'.
  4. [Section 5.6.1] The abbreviation 'RSME' appears where 'RMSE' is intended.
  5. [Section 5.3] The relationship between the road split in Table 1 and the randomly selected 20% temporal testing period should be described more precisely, since the current wording makes it unclear whether the temporal split is applied only to testing roads or to all roads.
  6. [Figure 7 and surrounding text] The label 'unseen sensors' would be more precise as 'held-out sensors', since these roads still have sensors and ground-truth flow.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the estimation is a supervised regression from static map images and speed data to flow, with independent held-out evaluation.

full rationale

The paper's central derivation is not circular. The proposed function Phi maps a spatiotemporal graph (node attributes: speed, lanes, coordinates) plus three GOMS map images (OSM, sensor distribution, population density) to estimated flow; ground-truth flow appears only as the regression target in Eq. (1) and in the loss/evaluation metrics (Eqs. 14-17), never as an input feature. The sensor distribution map is constructed from sensor coordinates ('The sensor locations colored with red are drawn using the longitude and latitude information in the UTD19 and PeMS datasets'), so it encodes location, not flow values, and does not by construction determine the output. The method is evaluated on road segments held out from training against non-trivial baselines (SA, LSTM, GCMC, GATv2), and the ablation study varies input maps and network components, providing independent empirical support. Self-citations (e.g., Loder et al. 2019 for UTD19 data, MFD work by Ambühl/Menendez) are data sources or background, not load-bearing derivations. The main caveat is a generalization gap rather than circularity: every evaluated target is a sensor location whose own marker appears in the sensor distribution map and whose speed is an input, while the motivating scenario of truly unmonitored roads would lack both; this is a train/serve mismatch that should be assessed as a correctness/generality risk, not as the output reducing to the input.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or theoretical entities. Its central claim rests on domain assumptions about the predictive power of static maps and on several architecture hyperparameters chosen by hand. The main unexamined assumption is the consistency of the sensor distribution map between training (target sensor marked) and deployment (target sensor absent).

free parameters (4)
  • Time window M = 12 (one hour at 5-minute intervals)
    Historical window length for graph temporal attention; chosen by hand without sensitivity analysis across cities.
  • Embedding dimension = 512
    Embedding size for graph and image decoders; fixed without reported tuning.
  • Image zoom levels = 15, 16, 17
    Zoom levels for OSM and sensor distribution maps; selected ad hoc, affecting the spatial context captured.
  • Train/testing road split = varies by city (e.g., 422 total / 347 train / 75 test for Los Angeles)
    The split of roads into training and testing sets is a design choice that influences reported errors; sensitivity is checked only for two cities.
assumptions (4)
  • domain assumption Speed on a link is correlated with flow on the same link.
    The model uses observed speed as the primary temporal input to estimate flow (Section 3.1); the correlation is learned from data and not guaranteed to be stable across cities or congestion regimes.
  • domain assumption Static GOMS maps contain information predictive of traffic flow.
    The whole method relies on OSM, population density, and sensor distribution maps being informative for flow on unobserved roads (Sections 1 and 5.5).
  • domain assumption Training-road patterns transfer to testing roads within the same city.
    Evaluation withholds entire roads; the method assumes the learned speed-flow-map relationship generalizes to unseen roads in that city (Section 5.4).
  • domain assumption Ground-truth flow is available on all road segments for evaluation.
    The paper states that all road segments have flow data for evaluation purposes (Section 5.1.1), which is an idealized version of the deployment scenario where unmonitored roads lack ground truth.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Network-Wide Traffic Flow Estimation Across Multiple Cities with Global Open Multi-Source Data: A Large-Scale Case Study in Europe and North America." pith.science (2026). https://pith.science/paper/URFSZNMH

@misc{pith2026250203798,
  author       = {Pith},
  title        = {Pith review of: Network-Wide Traffic Flow Estimation Across Multiple Cities with Global Open Multi-Source Data: A Large-Scale Case Study in Europe and North America},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/URFSZNMH}},
  note         = {Machine review of arXiv:2502.03798}
}
read the original abstract

Network-wide traffic flow, which captures dynamic traffic volume on each link of a general network, is fundamental to smart mobility applications. However, the observed traffic flow from sensors is usually limited across the entire network due to the associated high installation and maintenance costs. To address this issue, existing research uses various supplementary data sources to compensate for insufficient sensor coverage and estimate the unobserved traffic flow. Although these studies have shown promising results, the inconsistent availability and quality of supplementary data across cities make their methods typically face a trade-off challenge between accuracy and generality. In this research, we first time advocate using the Global Open Multi-Source (GOMS) data within an advanced deep learning framework to break the trade-off. The GOMS data primarily encompass geographical and demographic information, including road topology, building footprints, and population density, which can be consistently collected across cities. More importantly, these GOMS data are either causes or consequences of transportation activities, thereby creating opportunities for accurate network-wide flow estimation. Furthermore, we use map images to represent GOMS data, instead of traditional tabular formats, to capture richer and more comprehensive geographical and demographic information. To address multi-source data fusion, we develop an attention-based graph neural network that effectively extracts and synthesizes information from GOMS maps while simultaneously capturing spatiotemporal traffic dynamics from observed traffic data. A large-scale case study across 15 cities in Europe and North America was conducted. The results demonstrate stable and satisfactory estimation accuracy across these cities, which suggests that the trade-off challenge can be successfully addressed using our approach.

Figures

Figures reproduced from arXiv: 2502.03798 by the authors.

Figure 1
Figure 1. An example of the proposed NTFE framework. Section 3.1, and then define the problem in Section 3.2. 3.1. Data description The utilized multi-source data are three-fold, public traffic data, network data, and GOMS maps, which are detailed as follows: Traffic speed data In the existing literature, traffic speed data has been extensively utilized in NTFE due to the correlation between flow and speed, which can provide … view at source ↗
Figure 2
Figure 2. An overview of the proposed attention-based graph neural network with GOMS maps. 4.1. Image processor In this section, four modules in the image processor named, image encoder, image attention module, dense module, and image decoder, will be elaborated separately. 4.1.1. Image encoder The OSM, sensor distribution map, and population density map are pre-encoded by three individual image encoders. We employ a convolut… view at source ↗
Figure 3
Figure 3. The pipeline of the TCAB and the DCB. and 𝐼 𝑝,𝑣 𝑖 ) will react with feature maps from the query blocks (𝐼 𝑜,𝑞 𝑖 , 𝐼 𝑙,𝑞 𝑖 and 𝐼 𝑝,𝑞 𝑖 ). The formulation of the TCAB is shown in Equations 5 and 6:  𝑡 𝑖 = [ 𝐼 𝑜,𝑞 𝑖 , 𝐼𝑙,𝑞 𝑖 , 𝐼𝑝,𝑞 𝑖 ]𝑇 ○※ [ 𝐼 𝑜,𝑣 𝑖 , 𝐼𝑙,𝑣 𝑖 , 𝐼𝑝,𝑣 𝑖 ] = ⎡ ⎢ ⎢ ⎢ ⎢ ⎣  𝑜,𝑜 𝑡 ( 𝐼 𝑜,𝑞 𝑖 𝐼 𝑜,𝑣 𝑖 )  𝑜,𝑙 𝑡 ( 𝐼 𝑜,𝑞 𝑖 𝐼 𝑙,𝑣 𝑖 )  𝑜,𝑝 𝑡 ( 𝐼 𝑜,𝑞 𝑖 𝐼 𝑝,𝑣 𝑖 )  𝑙,𝑜 𝑡 ( 𝐼 𝑙,𝑞 𝑖 𝐼 𝑜,𝑣 𝑖 )  𝑙,𝑙 𝑡 ( 𝐼 𝑙,𝑞 𝑖 𝐼 𝑙,𝑣 𝑖… view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Snapshots of road segments with the placement of fixed sensors. world. Moreover, it is worth noting that the OSM and sensor distribution map are 3-channel RGB images, while the density maps are 1-channel gray-scale images. The gray-scale value indicates the population …
Figure 5
Figure 5. Figure 5: Snapshots from GOMS maps in Manchester. 5(a)-5(c) are OSMs with zoom level of 15-17. 5(d)-5(f) are sensor distribution maps with zoom levels of 15-17.5(g)-5(i) are population density maps with zoom levels of 15-17. training data is not enough resulting in overfitting i…
Figure 6
Figure 6. Figure 6: Boxplots of the estimation errors in all cities. The units for RMSE and MAE are veh/hour/lane [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]
Figure 7
Figure 7. Figure 7: Variation of average daily traffic flow of unseen sensors from the proposed method and the ground truth. Z.H, Z.Z, M.M, W.M: Preprint submitted to Elsevier Page 17 of 25 [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: Correlation of the estimated flow with the ground truth. If points are close to the red lines, the estimations are accurate. Z.H, Z.Z, M.M, W.M: Preprint submitted to Elsevier Page 18 of 25 [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: Results of ablation studies on data sources within all cities. The black lines represent the standard deviation. Importantly, the proposed estimation method is flexible to add or replace some data sources. With the development of sensor technologies, it is foreseeable …
Figure 10
Figure 10. Figure 10: The boxplots of estimation errors of ablated components within all cities. neural network. The maximal, minimal and median errors are reduced when we incorporate the TCAB or DCB. The estimation error is further reduced when we incorporate all blocks in the neural netw…
Figure 11
Figure 11. Figure 11: The estimation errors of our method on different training sets in Manchester and Oakland. The X-axis denotes the proportion of selected nodes in the original training set. The left and right Y-axes mean errors in Manchester and Oakland separately. 5.6.2. Sensitivity a…
Figure 12
Figure 12. Figure 12: The boxplots of estimation errors with different training and testing sets in Manchester and Oakland. In each subplot, the red and green dashed lines represent the baseline errors in Manchester and Oakland in [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Network-Wide Traffic Volume Estimation from Speed Profiles using a Spatio-Temporal Graph Neural Network with Directed Spatial Attention

    cs.LG 2025-12 conditional novelty 6.0 of 10

    An inductive spatio-temporal graph network estimates network-wide daily traffic volumes from speed profiles and road attributes, without volume data at inference.

  2. Capacity-Aware Deep Learning for Generalizable Traffic Volume Estimation Across Links and Cities

    cs.LG 2026-07 conditional novelty 5.0 of 10

    Factoring predicted traffic volume into a learned link capacity and a regime-aware utilization ratio improves hourly volume estimates on unseen links and across cities.

Reference graph

Works this paper leans on

63 extracted references · 59 canonical work pages · cited by 2 Pith papers

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  2. [2]

    Adam-Poupart, A., Brand, A., Fournier, M., Jerrett, M., and Smargiassi, A. (2014). Spatiotemporal modeling of ozone levels in quebec ( C anada): a comparison of kriging, land-use regression ( LUR) , and combined bayesian maximum entropy-LUR approaches. Environmental Health Perspectives , 122(9):970--976

  3. [3]

    Ambühl, L., Loder, A., Bliemer, M. C. J., Menendez, M., and Axhausen, K. W. (2018). Introducing a re-sampling methodology for the estimation of empirical macroscopic fundamental diagrams. Transportation Research Record , 2672(20):239--248

  4. [4]

    and Menendez, M

    Ambühl, L. and Menendez, M. (2016). Data fusion algorithm for macroscopic fundamental diagram estimation. Transportation Research Part C: Emerging Technologies , 71:184--197

  5. [5]

    Anuar, K., Habtemichael, F., and Cetin, M. (2015). Estimating traffic flow rate on freeways from probe vehicle data and fundamental diagram. In 2015 IEEE 18th international conference on intelligent transportation systems , pages 2921--2926. IEEE

  6. [6]

    and Yamagata, Y

    Bagan, H. and Yamagata, Y. (2015). Analysis of urban growth and estimating population density using satellite images of nighttime lights and land-use and population data. GIScience & Remote Sensing , 52(6):765--780

  7. [7]

    Brody, S., Alon, U., and Yahav, E. (2022). How attentive are graph attention networks? In International Conference on Learning Representations

  8. [8]

    M., Benitez, F

    Caceres, N., Romero, L. M., Benitez, F. G., and del Castillo, J. M. (2012). Traffic flow estimation models using cellular phone data. IEEE Transactions on Intelligent Transportation Systems , 13(3):1430--1441

Show all 63 references
  1. [9]

    Chen, X., Chen, Y., Saunier, N., and Sun, L. (2021). Scalable low-rank tensor learning for spatiotemporal traffic data imputation. Transportation Research Part C: Emerging Technologies , 129:103226

  2. [10]

    Chen, X., He, Z., and Sun, L. (2019). A bayesian tensor decomposition approach for spatiotemporal traffic data imputation. Transportation Research Part C: Emerging Technologies , 98:73--84

  3. [11]

    Chen, X., Lei, M., Saunier, N., and Sun, L. (2022). Low-rank autoregressive tensor completion for spatiotemporal traffic data imputation. IEEE Transactions on Intelligent Transportation Systems , 23(8):12301--12310

  4. [12]

    Chen, X., Yang, J., and Sun, L. (2020a). A nonconvex low-rank tensor completion model for spatiotemporal traffic data imputation. Transportation Research Part C: Emerging Technologies , 117:102673

  5. [13]

    Chen, Y., Lv, Y., and Wang, F.-Y. (2020b). Traffic flow imputation using parallel data and generative adversarial networks. IEEE Transactions on Intelligent Transportation Systems , 21(4):1624--1630

  6. [14]

    Cleland, J. (1996). Demographic data collection in less developed countries 1946–1996. Population Studies , 50(3):433--450

  7. [15]

    Cui, Z., Henrickson, K., Ke, R., and Wang, Y. (2020). Traffic graph convolutional recurrent neural network: A deep learning framework for network-scale traffic learning and forecasting. IEEE Transactions on Intelligent Transportation Systems , 21(11):4883--4894

  8. [16]

    Daganzo, C. F. (1997). Fundamentals of transportation and traffic operations . Emerald Group Publishing Limited

  9. [17]

    Daganzo, C. F. and Geroliminis, N. (2008). An analytical approximation for the macroscopic fundamental diagram of urban traffic. Transportation Research Part B: Methodological , 42(9):771--781

  10. [18]

    and Menendez, M

    Dakic, I. and Menendez, M. (2018). On the use of lagrangian observations from public transport and probe vehicles to estimate car space-mean speeds in bi-modal urban networks. Transportation Research Part C: Emerging Technologies , 91:317--334

  11. [19]

    Dunn, R. J. H., Willett, K. M., Parker, D. E., and Mitchell, L. (2016). Expanding hadisd: quality-controlled, sub-daily station data from 1931. Geoscientific Instrumentation, Methods and Data Systems , 5(2):473--491

  12. [20]

    Fan, C., Yang, Y., and Mostafavi, A. (2024). Neural embeddings of urban big data reveal spatial structures in cities. Humanities and Social Sciences Communications , 11(1):409

  13. [21]

    Fedorov, A., Nikolskaia, K., Ivanov, S., Shepelev, V., and Minbaleev, A. (2019). Traffic flow estimation with data from a video surveillance camera. Journal of Big Data , 6(1):73

  14. [22]

    He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , pages 770--778

  15. [23]

    A., Chawla, S., Alizadeh, M., Balakrishnan, H., and Madden, S

    He, S., Sadeghi, M. A., Chawla, S., Alizadeh, M., Balakrishnan, H., and Madden, S. (2021). Inferring high-resolution traffic accident risk maps based on satellite imagery and gps trajectories. In 2021 IEEE/CVF International Conference on Computer Vision (ICCV) , pages 11957--11965

  16. [24]

    and Schmidhuber, J

    Hochreiter, S. and Schmidhuber, J. (1997). Long short-term memory. Neural Computation , 9(8):1735--1780

  17. [25]

    H., Chen, G

    Kaack, L. H., Chen, G. H., and Morgan, M. G. (2019). Truck traffic monitoring with satellite images. In Proceedings of the 2nd ACM SIGCAS Conference on Computing and Sustainable Societies , page 155–164. Association for Computing Machinery

  18. [26]

    Kerner, B. S. (2009). Introduction to modern traffic flow theory and control , volume 700. Springer

  19. [27]

    Kingma, D. P. and Ba, J. (2017). Adam: A method for stochastic optimization

  20. [28]

    Lam, W. H. K., Chan, K. S., and Shi, J. W. Z. (2002). A traffic flow simulator for short-term travel time forecasting. Journal of Advanced Transportation , 36(3):265--291

  21. [29]

    Lam, W. H. K. and Yin, Y. (2001). An activity-based time-dependent traffic assignment model. Transportation Research Part B , 35(6):549--574

  22. [30]

    Laraki, M., De Nunzio, G., and Othman, B. (2022). A large-scale and data-based road traffic flow estimation method leveraging topography information and population statistics. In 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC) , pages 2343--2349

  23. [31]

    Li, J., Boonaert, J., Doniec, A., and Lozenguez, G. (2021). Multi-models machine learning methods for traffic flow estimation from floating car data. Transportation Research Part C: Emerging Technologies , 132:103389

  24. [32]

    Li, L., Li, Y., and Li, Z. (2013). Efficient missing data imputing for traffic flow by considering temporal and spatial dependence. Transportation Research Part C: Emerging Technologies , 34:108--120

  25. [33]

    Liu, Z., Liu, Y., Meng, Q., and Cheng, Q. (2019). A tailored machine learning approach for urban transport network flow estimation. Transportation Research Part C: Emerging Technologies , 108:130--150

  26. [34]

    Loder, A., Amb \"u hl, L., Menendez, M., and Axhausen, K. W. (2019). Understanding traffic capacity of urban networks. Scientific Reports , 9(1):16283

  27. [35]

    Mahajan, V., Cantelmo, G., Rothfeld, R., and Antoniou, C. (2023). Predicting network flows from speeds using open data and transfer learning. IET Intelligent Transport Systems , 17(4):804--824

  28. [36]

    L., Laña, I., Villar-Rodriguez, E., and Ser, J

    Manibardo, E. L., Laña, I., Villar-Rodriguez, E., and Ser, J. D. (2023). A graph-based methodology for the sensorless estimation of road traffic profiles. IEEE Transactions on Intelligent Transportation Systems , 24(8):8701--8715

  29. [37]

    Meng, C., Yi, X., Su, L., Gao, J., and Zheng, Y. (2017). City-wide traffic volume inference with loop detector data and taxi trajectories. In Proceedings of the 25th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems , SIGSPATIAL '17. Associa...

  30. [38]

    and Head, L

    Mirchandani, P. and Head, L. (2001). A real-time traffic signal control system: architecture, algorithms, and analysis. Transportation Research Part C: Emerging Technologies , 9(6):415--432

  31. [39]

    Ni, D. (2016). Traffic Flow Theory . Butterworth-Heinemann

  32. [40]

    Nie, T., Qin, G., Wang, Y., and Sun, J. (2023). Towards better traffic volume estimation: Jointly addressing the underdetermination and nonequilibrium problems with correlation-adaptive GNN s. Transportation Research Part C: Emerging Technologies , 157:104402

  33. [41]

    Othman, B., De Nunzio, G., Laraki, M., and Sabiron, G. (2022). A novel approach to traffic flow estimation based on floating car data and road topography: Experimental validation in lyon, F rance. In 2022 IEEE 25th International Conference on Intelligent Transportation Systems...

  34. [42]

    and Jackson, I

    Pavlyuk, D. and Jackson, I. (2022). Potential of vision-enhanced floating car data for urban traffic estimation. Transportation Research Procedia , 62:366--373

  35. [43]

    Pelletier, M.-P., Tr \'e panier, M., and Morency, C. (2011). Smart card data use in public transit: A literature review. Transportation Research Part C: Emerging Technologies , 19(4):557--568

  36. [44]

    Pendyala, R. M. and Pas, E. I. (2000). Multi-day and multi-period data for travel demand analysis and modeling. In Transportation Research Board Annual Meeting , page 22

  37. [45]

    Pun, L., Zhao, P., and Liu, X. (2019). A multiple regression approach for traffic flow estimation. IEEE Access , 7:35998--36009

  38. [46]

    Qu, L., Li, L., Zhang, Y., and Hu, J. (2009). PPCA -based missing data imputation for traffic flow volume: A systematical approach. IEEE Transactions on Intelligent Transportation Systems , 10(3):512--522

  39. [47]

    Qu, L., Zhang, Y., Hu, J., Jia, L., and Li, L. (2008). A BPCA based missing value imputing method for traffic flow volume data. In 2008 IEEE Intelligent Vehicles Symposium , pages 985--990

  40. [48]

    Rianne van den Berg, Thomas N. Kipf, M. W. (2017). Graph convolutional matrix completion. In Proceedings of the 25th ACM SIGKDD international conference on Knowledge discovery and data mining

  41. [49]

    Ross, P. (1988). Traffic dynamics. Transportation Research Part B: Methodological , 22(6):421--435

  42. [50]

    V., Sadabadi, K

    Seku a, P., Laan, Z. V., Sadabadi, K. F., Kania, K., and Zahedian, S. (2021). Transferability of a machine learning-based model of hourly traffic volume estimation--- F lorida and N ew H ampshire case study. Journal of Advanced Transportation , 2021:9944918

  43. [51]

    and Chen, L

    Shao, W. and Chen, L. (2018). License plate recognition data-based traffic volume estimation using collaborative tensor decomposition. IEEE Transactions on Intelligent Transportation Systems , 19(11):3439--3448

  44. [52]

    Steininger, M., Kobs, K., Zehe, A., Lautenschlager, F., Becker, M., and Hotho, A. (2020). MapLUR : Exploring a new paradigm for estimating air pollution using deep learning on map images. ACM Transactions on Spatial Algorithms and Systems , 6(3)

  45. [53]

    G., Liu, X., Zhang, A., Gros, A., Li, N., Yetman, G., Kilic, T., Murray, S., Blankespoor, B., Prydz, E

    Tiecke, T. G., Liu, X., Zhang, A., Gros, A., Li, N., Yetman, G., Kilic, T., Murray, S., Blankespoor, B., Prydz, E. B., and Dang, H.-A. H. (2017). Mapping the world population one building at a time

  46. [54]

    Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., and Bengio, Y. (2018). Graph attention networks. In International Conference on Learning Representations

  47. [55]

    Vuchic, V. R. (2007). Urban transit systems and technology . John Wiley & Sons

  48. [56]

    Wang, P., Lai, J., Huang, Z., Tan, Q., and Lin, T. (2021). Estimating traffic flow in large road networks based on multi-source traffic data. IEEE Transactions on Intelligent Transportation Systems , 22(9):5672--5683

  49. [57]

    Xing, J. (2022). Estimating traffic volumes in an urban network based on taxi gps and limited flow data using machine learning techniques. In 2022 IEEE 25th International Conference on Intelligent Transportation Systems (ITSC) , pages 20--25

  50. [58]

    C., Markovic, N., and Phillips, J

    Yi, Z., Liu, X. C., Markovic, N., and Phillips, J. (2021). Inferencing hourly traffic volume using data-driven machine learning and graph theory. Computers, Environment and Urban Systems , 85:101548

  51. [59]

    Zhan, X., Li, R., and Ukkusuri, S. V. (2020). Link-based traffic state estimation and prediction for arterial networks using license-plate recognition data. Transportation Research Part C: Emerging Technologies , 117:102660

  52. [60]

    Zhan, X., Zheng, Y., Yi, X., and Ukkusuri, S. V. (2017). Citywide traffic volume estimation using trajectory data. IEEE Transactions on Knowledge and Data Engineering , 29(2):272--285

  53. [61]

    Zhang, Y., Cheng, Q., Liu, Y., and Liu, Z. (2023). Full-scale spatio-temporal traffic flow estimation for city-wide networks: a transfer learning based approach. Transportmetrica B: Transport Dynamics , 11(1):869--895

  54. [62]

    Zhang, Z., Li, M., Lin, X., and Wang, Y. (2020). Network-wide traffic flow estimation with insufficient volume detection and crowdsourcing data. Transportation Research Part C: Emerging Technologies , 121:102870

  55. [63]

    Zhong, M., Lingras, P., and Sharma, S. (2004). Estimation of missing traffic counts using factor, genetic, neural, and regression techniques. Transportation Research Part C: Emerging Technologies , 12(2):139--166

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.