Pith. sign in

REVIEW 5 major objections 5 minor 35 references

Pairwise Spatiotemporal Partial Trajectory Matching for Co-movement Analysis

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read A Siamese network classifies co-walking from GPS-only trajectory images at F1 0.73.

desk verdict A genuinely interesting way to visualize and compare GPS trajectories, but the evaluation as reported does not support the headline 0.73 F1 claim. read the letter →

arxiv 2412.02879 v1 pith:7ZCCKNLR submitted 2024-12-03 cs.CV

classification cs.CV
keywords co-walkingdetectiontrajectorymatchingSiamesenetworkGPStrajectoriesspatiotemporaldataimagelayerspartialroutinebehavioranalysis
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 proposes classifying whether two people walked together on a given day by turning each person's GPS track into a series of time-windowed images, with color encoding time along the route, and feeding paired images to a Siamese network. The authors show this image-based pipeline reaches an F1 score up to 0.73 and MCC 0.45 on a 126-participant five-week co-walking study, beating LSTM and GRU baselines trained on raw coordinates. The key methodological finding is that most of the gain comes from a spatial-overlap check before classification, while additional filtering and cropping steps do not help. The broader claim is that layered trajectory images provide an interpretable way to match partial trajectories and to expose when, how often, and how long a pair's routines coincide.

What carries the argument

The central object is the layered trajectory image: raw latitude/longitude/time rows are split into 1, 5, 24, or 48 time windows per day, each window is drawn on a shared canvas with a color gradient showing time progression, and each person gets one image per layer. The pipeline then localizes trajectories with Selective Search bounding boxes, rejects layer pairs whose boxes do not spatially overlap, and sends only overlapping pairs to SigNet, a Siamese convolutional network with contrastive loss, to output a similarity distance for the layer.

What would settle it

Run the fine-tuned 48-layer overlap-only model under leave-one-pair-out cross-validation, where no pair appears in both training and test sets; if the F1 score falls from 0.73 to near the LSTM's 0.58, the reported gain comes from route memorization rather than co-walking detection.

Watch

Extended reading notes

Core claim

In the paper's own terms, the discovery is that pairwise spatiotemporal partial trajectory matching can be recast as an image-similarity problem: slice a day into layers, draw each person's trajectory as a colored line on a shared canvas, check whether the two trajectory bounding boxes overlap, and let a Siamese convolutional network decide whether the pair was co-walking during that interval. With localization and overlap checking but without filtering or cropping, the fine-tuned SigNet model reaches F1 0.73, accuracy 0.74, and MCC 0.45 on 48 layers, and the pre-trained model reaches F1 0.73 on 24 layers; both outperform sequence models. The authors position this as the first time paired walking instances have been classified with notable precision from GPS data alone.

Load-bearing premise

The training labels assume that a day-level self-report about whether a pair walked together can be attached to each individual time-window image pair, and that the random 5-fold split does not let the same pair's other days leak into training.

Editorial extensions

If this is right

  • With 24-layer images and only localization plus overlap checking, the pre-trained SigNet reaches F1 0.73 and MCC 0.41; the fine-tuned model reaches F1 0.73 and MCC 0.45 at 48 layers.
  • Both configurations beat the LSTM (F1 0.58, MCC 0.16) and GRU (F1 0.54, MCC 0.06) baselines in the paper's experiments.
  • Filtering low-information layers and cropping to bounding boxes reduce performance, so the simplest pipeline is recommended.
  • The layer index of a positive match gives the time window of the shared behavior, enabling routine-pattern analysis of frequency, timing, and duration.
  • The method is slower by orders of magnitude than sequence models, so it fits offline analysis where accuracy and interpretability matter more than speed.

Reading between the lines

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

  • Because the ablation attributes most of the improvement to the bounding-box overlap check, a fair skeptical test is whether a Siamese network adds anything beyond 'same place, same time' filtering; the paper's numbers do not resolve that question.
  • The same layered-image recipe could be applied to other paired behaviors, such as co-commuting or joint errands, with only the fine-tuning dataset changed.
  • The visual output could support privacy-conscious applications: an analyst or participant can see exactly which time interval triggered a positive match.
  • A direct baseline that thresholds GPS distance between paired points would isolate the contribution of image features; the paper does not report such a baseline.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 5 minor

Summary. The paper proposes an image-based pipeline for pairwise spatiotemporal partial trajectory matching. It converts raw GPS trajectories of two individuals into layered trajectory images over time windows, filters out low-quality images, localizes and crops trajectories, checks spatial overlap, and uses a Siamese SigNet network to classify whether a layer pair corresponds to co-walking. The method is evaluated on a 5-week pair-walking dataset with 1,472 person-day instances and compared with LSTM and GRU baselines. The authors report F1 scores up to 0.73, claim the method outperforms established baselines, and present an ablation study plus a qualitative routine-pattern analysis.

Significance. If the empirical claims were supported, the paper would offer a novel, interpretable alternative for co-movement detection from GPS data, with the useful property of identifying partial time windows of shared activity. The paper is clearly written, and the ablation study is a sensible attempt to decompose the pipeline. However, the main contribution—the demonstrated superiority over baselines—rests on an evaluation protocol with likely data leakage, an underspecified label-construction step, and selective reporting of an intermediate configuration. The current evidence is therefore not sufficient to establish the claimed performance.

major comments (5)
  1. [Section 4.2.3 / Section 4.1] The 5-fold cross-validation is described without any grouping by participant pair or individual. The dataset contains 94 individuals (47 pairs) and 1,472 person-day instances over 38 days; a random instance-level split will place other days from the same pair in the training set. Because walking routes are habitual, the Siamese network can then recognize a pair's usual geography rather than detect an actual co-walk, inflating the reported F1. The authors should explicitly state how folds were constructed; if folds are not group-based, the evaluation must be re-run with pair- or person-grouped folds.
  2. [Section 4.1 / Section 4.2.3] The construction of layer-level training labels for the fine-tuned Siamese network is never specified. The paper states that daily self-reports only indicate whether the pair walked together at some point during the day, not when or where. It is therefore unclear how each image-layer pair is assigned a positive or negative label for contrastive training. If all layers of a positive day are labeled positive, the training set contains systematic false positives; if only overlapping layers are used, the rule is not given. This omission is load-bearing because it determines the learned similarity function.
  3. [Section 4.5.2 / Table 2 / Abstract] The headline result "F1-score up to 0.73" is taken from the Overlap-only ablation configuration (FT, 48 layers), not from the complete pipeline: Table 1 reports F1=0.64 for the entire method with 48 layers. The conclusion repeats "F1 score of 0.73" and "significantly outperforming benchmark methods" based on this intermediate configuration that was selected after inspecting the same cross-validation folds. This selective reporting overstates the performance of the proposed pipeline as defined in Section 3.
  4. [Section 4.2.2 / Section 5.1] The pixel filtering thresholds (25th percentile for 1-layer, 50th for others) and the choice of layer count are determined empirically on the same dataset, and the best configuration is selected from the same cross-validation folds used to report final numbers. Without nested or held-out validation, the reported metrics reflect selection bias. The authors themselves acknowledge sensitivity to parameter choices in Section 5.1, which further supports the need for a proper validation scheme.
  5. [Section 4.3] The benchmark comparison is under-specified and possibly unfair. The LSTM and GRU hyperparameters (number of hidden units, sequence length, input representation) are not given; the preprocessing step that "retain[s] only those with matching lengths" is not described; and it is unclear how missing data and the same 5-fold split are handled. Without these details, the claimed out-performance of LSTM/GRU cannot be reproduced or assessed.
minor comments (5)
  1. [Throughout] The paper contains several typos and formatting issues, including "SignNet" for "SigNet" in Section 4.5.2, "P Tindicating" in the Figure 3 caption, and "V A" for "VA" in the author affiliations.
  2. [Equation (1)] Equation (1) uses w1 and w2 for the two subnetworks but states they share weights; clarify and use a single weight vector.
  3. [Section 4.4] Section 4.4 refers to "average weighted results" but does not define the weighting scheme; please specify whether these are macro/micro averages.
  4. [Section 4.6] The routine pattern analysis in Section 4.6 is qualitative and presents only a single illustrative example; a quantitative evaluation or multiple examples would strengthen the claim.
  5. [Section 4.2.3] Section 4.2.3 states the models are trained for 50 epochs but does not report the number of training samples per fold or the batch size; including these details would aid reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported co-walking classification is an empirical, externally benchmarked evaluation rather than a result forced by construction or by a self-citation chain.

full rationale

The paper's central claim is an empirical performance comparison: image-layer generation plus a Siamese network classifies co-walking days with F1 up to 0.73. The derivation chain is a data-processing and training pipeline, not an analytical derivation in which an output equation follows from an input by construction. The image-generation, overlap-check, and Siamese-classification steps are each defined independently of the target F1 metric. The preprocessing thresholds are selected from pixel-count statistics and described as 'empirically determined' (Section 3.2) and set at percentiles 'based on empirical testing' (Section 4.2.2), but these are hyperparameters for filtering low-quality images, not fitted parameters that are then renamed as predictions. Similarly, the optimal configuration (24 or 48 layers, overlap without cropping/filtering) is chosen from the ablation results in Section 4.5.2 and then used to report the headline metrics; this is in-sample configuration selection, which can inflate reported numbers, but it does not make the F1 score equivalent to the fitting objective or to the inputs by definition. The SigNet architecture is cited from external work [32], and no load-bearing self-citation or author-imported uniqueness theorem appears anywhere in the paper. The acknowledged limitations in Section 5.1 concerning self-reported labels, missing data, and parameter sensitivity are data-quality and generalizability concerns rather than circularity. Possible data leakage from instance-level cross-validation is a validation-procedure risk, but it is not a circularity step under the required definition, since the reported score is still a measured classifier outcome rather than a mathematical restatement of the label-generation rule.

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

The method introduces no new physical entities. Its results rest on the private self-report ground truth, the unstated layer-level label mapping, hand-set filtering thresholds, and post-hoc configuration selection.

free parameters (3)
  • Pixel filtering threshold (percentile) = 25th percentile for 1-layer; 50th percentile for 5, 24, and 48-layer sets
    Empirically determined thresholds to discard low-quality image layers (Section 4.2.2).
  • Number of time layers = 24 and 48 selected as optimal
    Layer counts chosen after inspecting the same folds; 24 and 48 gave the best F1 during the ablation (Section 4.5.2).
  • Siamese similarity distance threshold = Not reported
    A threshold is applied to the Euclidean distance to classify pairs; its value is not stated, yet it determines the reported precision and recall.
assumptions (3)
  • domain assumption Self-reported daily surveys accurately indicate whether a pair co-walked that day.
    Ground truth labels come entirely from daily self-reports (Section 4.1); no independent verification is provided.
  • domain assumption If two trajectories' bounding boxes do not overlap in a time window, no co-walking occurred in that window.
    The overlap check (Section 3.4) hard-codes this rule, but co-walking individuals could be separated by GPS noise or bounding box inaccuracies.
  • domain assumption A SigNet model pre-trained on signature images transfers to GPS trajectory images.
    The authors assume signatures and trajectories share features such as shape and edges (Section 4.2.3).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pairwise Spatiotemporal Partial Trajectory Matching for Co-movement Analysis." pith.science (2026). https://pith.science/paper/7ZCCKNLR

@misc{pith2026241202879,
  author       = {Pith},
  title        = {Pith review of: Pairwise Spatiotemporal Partial Trajectory Matching for Co-movement Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7ZCCKNLR}},
  note         = {Machine review of arXiv:2412.02879}
}
read the original abstract

Spatiotemporal pairwise movement analysis involves identifying shared geographic-based behaviors between individuals within specific time frames. Traditionally, this task relies on sequence modeling and behavior analysis techniques applied to tabular or video-based data, but these methods often lack interpretability and struggle to capture partial matching. In this paper, we propose a novel method for pairwise spatiotemporal partial trajectory matching that transforms tabular spatiotemporal data into interpretable trajectory images based on specified time windows, allowing for partial trajectory analysis. This approach includes localization of trajectories, checking for spatial overlap, and pairwise matching using a Siamese Neural Network. We evaluate our method on a co-walking classification task, demonstrating its effectiveness in a novel co-behavior identification application. Our model surpasses established methods, achieving an F1-score up to 0.73. Additionally, we explore the method's utility for pair routine pattern analysis in real-world scenarios, providing insights into the frequency, timing, and duration of shared behaviors. This approach offers a powerful, interpretable framework for spatiotemporal behavior analysis, with potential applications in social behavior research, urban planning, and healthcare.

Figures

Figures reproduced from arXiv: 2412.02879 by the authors.

Figure 1
Figure 1. Proposed method pipeline, where l refers to the number of layers. The steps are labeled (a)-(h). The Siamese Network in (h) is the SigNet model, adapted from [32]. complex relationships or interactions among people from the image frames. In contrast, our method relies solely on a GPS sensor, a common feature in most smartphones and wearable devices such as smartwatches and rings. Our proposed method converts the ext… view at source ↗
Figure 2
Figure 2. Detailed view comparing a pair that co-walked on a day (row a) and a pair that did not co-walk on a day (row [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Positive class results (F1-score, precision, and recall) from co-walking classification. Each section of the radar [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: (a) Routine behavior patterns for two individuals in a pair across all days’ of data (Person 1: 23 days, Person [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Updated proposed method pipeline, where l refers to the number of layers. The steps are labeled (a)-(f). The Siamese Network in (f) is the SigNet model, adapted from [32]. This pipeline reflects the optimized configuration of our initial method based on empirical resul…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 32 canonical work pages

  1. [1]

    Analyzing movement predictability using human attributes and behavioral patterns

    Adir Solomon, Amit Livne, Gilad Katz, Bracha Shapira, and Lior Rokach. Analyzing movement predictability using human attributes and behavioral patterns. Computers, Environment and Urban Systems, 87:101596, May 2021. 14 A PREPRINT - DECEMBER 5, 2024

  2. [2]

    Personalized travel route recommendation using collaborative filtering based on GPS trajectories

    Ge Cui, Jun Luo, and Xin Wang. Personalized travel route recommendation using collaborative filtering based on GPS trajectories. International Journal of Digital Earth , 11(3):284–307, March 2018. Publisher: Taylor & Francis _eprint: https://doi.org/10.1080/17538947.2017.1326535

  3. [3]

    Duives, and Serge P

    Shi Sun, Cheng Sun, Dorine C. Duives, and Serge P. Hoogendoorn. Neural network model for predicting variation in walking dynamics of pedestrians in social groups. Transportation, 50(3):837–868, June 2023

  4. [4]

    Self-supervised human mobility learning for next location prediction and trajectory classification

    Fan Zhou, Yurou Dai, Qiang Gao, Pengyu Wang, and Ting Zhong. Self-supervised human mobility learning for next location prediction and trajectory classification. Knowledge-Based Systems, 228:107214, September 2021

  5. [5]

    Inferring Uncertain Trajectories from Partial Observations

    Prithu Banerjee, Sayan Ranu, and Sriram Raghavan. Inferring Uncertain Trajectories from Partial Observations, March 2016. arXiv:1603.07641

  6. [6]

    Show me how you move and I will tell you who you are

    Sébastien Gambs, Marc-Olivier Killijian, and Miguel Núñez Del Prado Cortez. Show me how you move and I will tell you who you are. In Proceedings of the 3rd ACM SIGSPATIAL International Workshop on Security and Privacy in GIS and LBS , pages 34–41, San Jose California, November 2010. ACM

  7. [7]

    Nascimento, Mário A

    Jacinto C. Nascimento, Mário A. T. Figueiredo, and Jorge S. Marques. Trajectory Classification Using Switched Dynamical Hidden Markov Models. IEEE Transactions on Image Processing, 19(5):1338–1348, May 2010

  8. [8]

    Predicting Destinations from Partial Trajecto- ries Using Recurrent Neural Network

    Yuki Endo, Kyosuke Nishida, Hiroyuki Toda, and Hiroshi Sawada. Predicting Destinations from Partial Trajecto- ries Using Recurrent Neural Network. In Jinho Kim, Kyuseok Shim, Longbing Cao, Jae-Gil Lee, Xuemin Lin, and Yang-Sae Moon, editors, Advances in Knowledge Discovery and Data Mining , pages 160–172, Cham, 2017. Springer International Publishing

Show all 35 references
  1. [9]

    Paschalidis, Hiroshi Taguchi, Satoko Sakajo, and Daniel Nikovski

    Athanasios Tsiligkaridis, Jing Zhang, Ioannis Ch. Paschalidis, Hiroshi Taguchi, Satoko Sakajo, and Daniel Nikovski. Context-Aware Destination and Time-To-Destination Prediction Using Machine learning. In 2022 IEEE International Smart Cities Conference (ISC2) , pages 1–7, Septe...

  2. [10]

    Efficient Trajectory Forecasting and Generation with Conditional Flow Matching, March 2024

    Sean Ye and Matthew Gombolay. Efficient Trajectory Forecasting and Generation with Conditional Flow Matching, March 2024. arXiv:2403.10809 version: 1

  3. [11]

    Early Destination Prediction with Spatio-temporal User Behavior Patterns

    Ryo Imai, Kota Tsubouchi, Tatsuya Konishi, and Masamichi Shimosaka. Early Destination Prediction with Spatio-temporal User Behavior Patterns. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 1(4):1–19, January 2018

  4. [12]

    The Places of Our Lives: Visiting Patterns and Automatic Labeling from Longitudinal Smartphone Data

    Trinh Minh Tri Do and Daniel Gatica-Perez. The Places of Our Lives: Visiting Patterns and Automatic Labeling from Longitudinal Smartphone Data. IEEE Transactions on Mobile Computing , 13(3):638–648, March 2014. Conference Name: IEEE Transactions on Mobile Computing

  5. [13]

    Modeling walking behavior of pedestrian groups with floor field cellular automaton approach

    Lu Li-Li, Ren Gang, Wang Wei, and Wang Yi. Modeling walking behavior of pedestrian groups with floor field cellular automaton approach. Chinese Physics B, 23(8):088901, June 2014

  6. [14]

    The Walking Behaviour of Pedestrian Social Groups and Its Impact on Crowd Dynamics

    Mehdi Moussaïd, Niriaska Perozo, Simon Garnier, Dirk Helbing, and Guy Theraulaz. The Walking Behaviour of Pedestrian Social Groups and Its Impact on Crowd Dynamics. PLOS ONE, 5(4):e10047, April 2010. Publisher: Public Library of Science

  7. [15]

    Potential for the dynamics of pedestrians in a socially interacting group

    Francesco Zanlungo, Tetsushi Ikeda, and Takayuki Kanda. Potential for the dynamics of pedestrians in a socially interacting group. Physical Review E, 89(1):012811, January 2014

  8. [16]

    M. S. Ryoo. Human activity prediction: Early recognition of ongoing activities from streaming videos. In 2011 International Conference on Computer Vision, pages 1036–1043, Barcelona, Spain, November 2011. IEEE

  9. [17]

    Recursive Social Behavior Graph for Trajectory Prediction

    Jianhua Sun, Qinhong Jiang, and Cewu Lu. Recursive Social Behavior Graph for Trajectory Prediction. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pages 657–666, Seattle, W A, USA, June 2020

  10. [18]

    STGAT: Modeling Spatial-Temporal Interactions for Human Trajectory Prediction

    Yingfan Huang, Huikun Bi, Zhaoxin Li, Tianlu Mao, and Zhaoqi Wang. STGAT: Modeling Spatial-Temporal Interactions for Human Trajectory Prediction. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 6271–6280, October 2019. ISSN: 2380-7504

  11. [19]

    Joint learning of social groups, individuals action and sub-group activities in videos

    Mahsa Ehsanpour, Alireza Abedin, Fatemeh Saleh, Javen Shi, Ian Reid, and Hamid Rezatofighi. Joint learning of social groups, individuals action and sub-group activities in videos. In Andrea Vedaldi, Horst Bischof, Thomas Brox, and Jan-Michael Frahm, editors, Computer Vision – ...

  12. [20]

    Urban Computing: Concepts, Methodologies, and Applications

    Yu Zheng, Licia Capra, Ouri Wolfson, and Hai Yang. Urban Computing: Concepts, Methodologies, and Applications. ACM Transactions on Intelligent Systems and Technology, 5(3):1–55, October 2014

  13. [21]

    Distributed Human Trajectory Sensing and Partial Similarity Queries

    Haotian Wang and Jie Gao. Distributed Human Trajectory Sensing and Partial Similarity Queries. In 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks (IPSN) , pages 253–264, Sydney, NSW, Australia, April 2020. 15 A PREPRINT - DECEMBER 5, 2024

  14. [22]

    Jeberson Retna Raj and T

    R. Jeberson Retna Raj and T. Sasipraba. Disaster management system based on GIS web services. In Recent Advances in Space Technology Services and Climate Change 2010 (RSTS & CC-2010) , pages 252–261, November 2010

  15. [23]

    Development of a Fully Automated, GPS Based Monitoring System for Disaster Prevention and Emergency Preparedness: PPMS+RT

    Jason Bond, Don Kim, Adam Chrzanowski, and Anna Szostak-Chrzanowski. Development of a Fully Automated, GPS Based Monitoring System for Disaster Prevention and Emergency Preparedness: PPMS+RT. Sensors, 7(7):1028–1046, July 2007. Number: 7 Publisher: Molecular Diversity Preserva...

  16. [24]

    Understanding User Behavior in Car Sharing Services Through The Lens of Mobility: Mixing Qualitative and Quantitative Studies

    Guang Wang, Harsh Rajkumar Vaish, Huijun Sun, Jianjun Wu, Shuai Wang, and Desheng Zhang. Understanding User Behavior in Car Sharing Services Through The Lens of Mobility: Mixing Qualitative and Quantitative Studies. Proceedings of the ACM on Interactive, Mobile, Wearable and U...

  17. [25]

    Buckee, Kenth Engø-Monsen, and C

    Amy Wesolowski, Caroline O. Buckee, Kenth Engø-Monsen, and C. J. E. Metcalf. Connecting Mobility to Infectious Diseases: The Promise and Limits of Mobile Phone Data. Journal of Infectious Diseases, 214(suppl 4):S414–S420, December 2016

  18. [26]

    GPS and Remote Sensing for Emergency Vehicle Navigation and Communication

    Farah Mahmood, Azzam ul Asar, and Adnan Mahmood. GPS and Remote Sensing for Emergency Vehicle Navigation and Communication. In 2006 International Conference on Advances in Space Technologies , pages 33–36, September 2006

  19. [27]

    Probabilistic Mining of Socio-Geographic Routines From Mobile Phone Data

    Katayoun Farrahi and Daniel Gatica-Perez. Probabilistic Mining of Socio-Geographic Routines From Mobile Phone Data. IEEE Journal of Selected Topics in Signal Processing , 4(4):746–755, August 2010

  20. [28]

    González

    Andrea Cuttone, Sune Lehmann, and Marta C. González. Understanding predictability and exploration in human mobility. EPJ Data Science, 7(1):1–17, December 2018. Number: 1 Publisher: SpringerOpen

  21. [29]

    Social-aware pedestrian trajectory prediction via states refinement lstm.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(5):2742– 2759, 2022

    Pu Zhang, Jianru Xue, Pengfei Zhang, Nanning Zheng, and Wanli Ouyang. Social-aware pedestrian trajectory prediction via states refinement lstm.IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(5):2742– 2759, 2022

  22. [30]

    Characterizing and predicting person-specific, day-to-day, fluctuations in walking behavior

    Guillaume Chevance, Dario Baretta, Matti Heino, Olga Perski, Merlijn Olthof, Predrag Klasnja, Eric Hekler, and Job Godino. Characterizing and predicting person-specific, day-to-day, fluctuations in walking behavior. PLOS ONE, 16(5):e0251659, May 2021. Publisher: Public Library...

  23. [31]

    Guiding Blind Pedestrians in Public Spaces by Understanding Walking Behavior of Nearby Pedestrians

    Seita Kayukawa, Tatsuya Ishihara, Hironobu Takagi, Shigeo Morishima, and Chieko Asakawa. Guiding Blind Pedestrians in Public Spaces by Understanding Walking Behavior of Nearby Pedestrians. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 4(3...

  24. [32]

    Ghosh, Josep Lladós, and Umapada Pal

    Sounak Dey, Anjan Dutta, Juan Ignacio Toledo, Suman K. Ghosh, Josep Lladós, and Umapada Pal. Signet: Convolutional siamese network for writer independent offline signature verification. ArXiv, abs/1707.02131, 2017

  25. [33]

    https://docs.opencv.org/4.x/d5/df0/group__ximgproc_ _segmentation.html

    Opencv image segmentation. https://docs.opencv.org/4.x/d5/df0/group__ximgproc_ _segmentation.html

  26. [34]

    Aware: mobile context instrumentation framework.Frontiers in ICT, 2:6, 2015

    Denzil Ferreira, Vassilis Kostakos, and Anind K Dey. Aware: mobile context instrumentation framework.Frontiers in ICT, 2:6, 2015

  27. [35]

    A density-based algorithm for discovering clusters in large spatial databases with noise

    Martin Ester, Hans-Peter Kriegel, Jörg Sander, and Xiaowei Xu. A density-based algorithm for discovering clusters in large spatial databases with noise. In Proceedings of the Second International Conference on Knowledge Discovery and Data Mining (KDD) , pages 226–231, 1996. 16...

Pith tools

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