Pith. sign in

REVIEW 5 major objections 7 minor 59 references

Pedestrian Trajectory Prediction Based on Social Interactions Learning With Random Weights

T0 review · 5 major / 7 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Multiplying graph attention by random edge weights lets a GAN learn pedestrian interactions without predefined rules, reaching 0.36/0.61 ADE/FDE on ETH/UCY.

desk verdict A modest empirical contribution whose central random-weight mechanism is under-specified and not properly ablated, but with enough solid experimental work to deserve a referee rather than a desk reject. read the letter →

arxiv 2501.07711 v1 pith:VESOMU5S submitted 2025-01-13 cs.CV cs.MM

Jiajia Xie , Sheng Zhang , Beihao Xia , Zhu Xiao , Hongbo Jiang , Siwang Zhou , Zheng Qin , Hongyang Chen This is my paper · ORCID
classification cs.CVcs.MM
keywords pedestriantrajectorypredictionsocialinteractionmodelinggraphrandomweightsgenerativeadversarialnetworkattentionGaussiannegativelog-likelihoodETH/UCYbenchmarkmulti-modalforecasting
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

Pedestrian trajectory prediction usually models social interaction by hand-coded rules such as Euclidean distance between walkers. The paper claims those rules are biased and incomplete: a target pedestrian may pay less attention to a nearby person behind them than to someone ahead they might collide with. To remove the rule, this work proposes DTGAN, a GAN that treats the observed scene at each time step as a graph whose edge weights are random values, multiplied into learned graph-attention coefficients before temporal decoding. With a Gaussian negative-log-likelihood task loss added to a Wasserstein adversarial objective, the best variant DTGAN-G reports average ADE/FDE of 0.36/0.61 meters on the ETH/UCY benchmark, ahead of listed baselines including Social-STGCNN at 0.44/0.75. The claim that matters: if random weights work, then interaction structure does not need to be specified in advance, and a network can discover it from data alone.

What carries the argument

The load-bearing mechanism is the random-weight attention multiplication of Eq. (3), $\hat\alpha^t_{ij} = \alpha^t_{ij} \cdot e^t_{ij}$, where $\alpha^t_{ij}$ is the standard GAT attention coefficient from Eq. (2) and $e^t_{ij}$ is a scalar edge weight attached to the pair of pedestrians $(i,j)$ at time $t$. The paper calls the matrix of these values a randomly sampled adjacency with zeros on the diagonal, and the related-work discussion adds that the weights are not fixed but evolve with training. That multiplication is the piece claimed to free the model from pre-defined interaction rules: none of the $e^t_{ij}$ values is a function of distance or direction, so the network cannot lean on a static rule and must learn which neighbors matter. The rest of the pipeline, including the temporal convolutional network, convolutional decoder, Wasserstein adversarial objective, and the chosen task loss, carries these randomly weighted node features into multi-modal trajectory predictions.

What would settle it

Train DTGAN-G identically on ETH/UCY with every edge multiplier fixed to 1 (a plain GAT) and compare ADE/FDE with the random-weight version. If the two are statistically indistinguishable, or if the plain-GAT version matches the reported 0.36/0.61, then the random weights are not carrying the interaction information; a complementary check is that the learned attention after training concentrates on pedestrians ahead or on collision course regardless of the sampled $e_{ij}$ values, which would mean the network learns interactions from coordinates and the random matrix is uninformative noise.

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that a randomly weighted adjacency matrix inside a graph attention network is enough to capture implicit social interactions among pedestrians. At each observed time step the model builds a graph $G_t = (V^t, A^t)$ whose nodes are pedestrians and whose adjacency entries satisfy $e^t_{ij} = z$ for $i \neq j$, where $z$ is a random value and the diagonal is zero. In Eq. (3) the learned GAT attention coefficient $\alpha^t_{ij}$ is multiplied by $e^t_{ij}$ to give $\hat\alpha^t_{ij}$, and the resulting node features pass through a temporal CNN, a convolutional decoder, and adversarial training. The paper contrasts this with Social-STGCNN's fixed reciprocal-distance weights, arguing that random weights remove the bias of rules like 'nearer means more important' and let the network assign importance adaptively. Tested on ETH and UCY with 8 observed frames and 12 predicted frames, the Gaussian negative-log-likelihood variant DTGAN-G reports average ADE/FDE of 0.36/0.61, which the authors read as evidence that the model understands pedestrian intent, especially in crowded UCY scenes.

Load-bearing premise

The load-bearing premise is that the random edge multipliers in Eq. (3), not the architecture, the adversarial loss, or the task loss, are what let the model capture implicit social interactions; the paper never isolates this premise because its comparisons swap random weights for Euclidean, all-ones, or arithmetic-progression matrices without ever running a plain GAT, and it never states the distribution of $z$ or whether $z$ is updated during training.

Editorial extensions

If this is right

  • If the random-weight mechanism works as claimed, graph-based interaction modeling in trajectory prediction no longer needs a hand-coded weighting function such as reciprocal distance, removing a source of scene-specific bias.
  • Adding a distributional task loss (Gaussian negative log-likelihood) to the adversarial objective yields the paper's largest measured gains: 16.7% on ADE and 39.3% on FDE relative to the plain DTGAN variant.
  • Because DTGAN treats each time step as a graph and decodes with TCN/CNN, adversarial training can be extended from flat sequences to graph-sequence data, which the paper frames as a new use of GANs.
  • On crowded UCY scenes with more social interactions, the reported gains over GAN-based baselines in ADE/FDE are larger than on ETH, which the paper reads as evidence that graph-structured node relationships help in complex crowds.
  • The robustness experiment across five random seeds reports small standard deviations, so the method's performance does not appear to depend on one lucky draw of the random matrix.

Reading between the lines

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

  • The random multipliers may be doing something more mundane than 'capturing implicit interactions': as fixed random noise they would act as a stochastic mask that regularizes attention, and as trainable parameters they would just be learned edge weights with random initialization; the paper's experiments do not distinguish these roles.
  • A direct way to test the paper's own story is to keep DTGAN-G but replace $e_{ij}$ with a learned function of relative position and velocity; if that model matches or beats random weights, then a data-driven rule, not randomness, is the active ingredient.
  • The same recipe, multiplying graph attention by random edge weights, could transfer to other GNN tasks that currently use predefined adjacency, such as multi-agent motion or traffic forecasting, but evidence would be needed that the benefit is not specific to the ETH/UCY setup.
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 / 7 minor

Summary. The manuscript proposes DTGAN, a GAN-based framework for pedestrian trajectory prediction in which random edge weights are introduced into graph attention. The generator processes graph sequences through spatial embedding, GAT, TCN, CNN, and a decoder, while the discriminator uses an SPE-LSTM-FC architecture. The paper explores three task-loss variants (MSE, Gaussian negative log-likelihood, and uniform negative log-likelihood), reports ADE/FDE and AMD/AMV results on ETH/UCY, and claims that DTGAN-G achieves 0.36/0.61 average ADE/FDE, outperforming all listed baselines because the random-weight mechanism automatically captures implicit social interactions without pre-defined rules.

Significance. If the mechanism and the quantitative claims were reliable, the paper would offer a simple modification to GAT-based trajectory predictors and a useful empirical study of task losses in adversarial training. The manuscript has strengths: comparisons against a broad set of baselines, a component ablation in Table V, a random-seed robustness study in Table IV, and qualitative analyses in Figures 3 and 4. However, the central random-weight mechanism is not specified in a way that can be implemented or tested, and several headline numbers in the abstract and Section IV-D are inconsistent with Table II. The scientific contribution therefore cannot currently be assessed as stated.

major comments (5)
  1. [Section III-A, Eq. (3), and Section II] The central mechanism is not defined. Section III-A states that e_{ij}=z for i≠j with z a random value and calls A_t a randomly sampled matrix, but no distribution, range, or sampling schedule is given. Section II says 'These weights are not fixed but are aligned with the model's training, allowing the representation of social interactions to evolve dynamically with training epochs', which is incompatible with fixed random values. If e_{ij} are fixed random constants, gradients do not flow into them, so Eq. (4) multiplies learned attention coefficients by static noise and the weights cannot 'learn' or evolve; if e_{ij} are trainable parameters, they are only randomly initialized and the claim that random weights eliminate pre-defined rules reduces to standard learned edge weights. The manuscript must specify the exact mechanism and experimentally distinguish frozen random, resampled, and trainable-random-initialization variants.
  2. [Section IV-F, Fig. 5] The comparison of weight-generation schemes does not isolate the random-weight contribution. The All Ones condition approximates standard GAT attention, but no standard GAT baseline without the multiplicative mask is reported, and no repeated runs or error bars are given for these four variants. Table IV reports random-seed variation only for the full DTGAN-G model. The gap between Random Weights (0.36/0.61) and All Ones (0.40/0.75) is therefore not shown to be statistically meaningful, so the claim that random weights specifically capture implicit social interactions is not supported.
  3. [Section IV-D and abstract] The claimed improvements of 16.7% for ADE and 39.3% for FDE do not correspond to any pairwise comparison in Table II. From DTGAN to DTGAN-G, the average changes are 0.42->0.36 (14.3%) and 0.85->0.61 (28.2%); against Social-STGCNN the changes are 18.2% and 18.7%. The authors should identify the reference model and correct the percentages.
  4. [Table II and Section IV-D] The claim that DTGAN-G gives 'the best performance among all baselines' rests on small margins over TPNMS and SocialDualCVAE (ADE 0.36 vs 0.38 and 0.38; FDE 0.61 vs 0.73 and 0.69), yet the headline results are reported without standard deviations, confidence intervals, or multiple-run statistics. Since the robustness analysis in Table IV exists, it should be extended to the final reported numbers and to the comparison variants in Fig. 5.
  5. [Section IV-H, Table V] The ablation study varies the GAT/GCN encoder, the SPE layer, and the temporal/decoder components, but it never removes or changes the random-weight mask. As a result, the component analysis does not test whether the random-weight term is responsible for any of the reported gains, which is the paper's central claim.
minor comments (7)
  1. [Section III-A] The sentence 'social interactions occur between more than two individuals' should read 'between pairs of individuals' or 'among more than two individuals'; as written, it conflicts with the zero diagonal and the off-diagonal definition of e_{ij}.
  2. [Section III-D, Eq. (13)] The notation TpY appears to be a misrendered product over t=1,...,Tp; please correct the product symbol and specify the parameters over which the argmin is taken.
  3. [Section III-D, Eq. (12)] The condition sqrt(xt_i^2 + yt_i^2) < r_hat uses absolute coordinates; it should use displacement from the predicted center of the circular distribution. Please clarify.
  4. [Section III-D] The sentence about WGAN cites 'Huang et al. [52]' but reference [52] is Lucic et al., 'Are GANs created equal?'; the citation should be corrected.
  5. [Section IV-B] Pre-training is mentioned as a separate stage, but no pre-training loss, number of epochs, or stopping criterion is provided; these details are needed for reproducibility.
  6. [Figure 6] The caption says each density curve represents a distribution of values, but the figure shows line plots of ADE/FDE versus random seeds; please align the caption with what is actually plotted.
  7. [Algorithm 1] The text 'Unique, Group, Len, Index and Append are specific function functions' contains a typo, and minPed is used in the algorithm but is not defined in the notation table. Please fix both.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: DTGAN is an empirical architecture paper whose random-weight mechanism and task-loss variants are evaluated on external benchmarks, not derived from the quantities they predict.

full rationale

The paper contains no step in which a claimed prediction or first-principles result is equivalent, by construction, to an input or to a fitted parameter. Equations (2)-(4) define a standard graph attention module whose learned coefficients are multiplied by externally sampled values e_ij; this is a forward-model architectural choice, and the paper does not define 'implicit social interactions' in terms of DTGAN's own outputs, so there is no self-definitional reduction. The ADE/FDE and AMD/AMV results in Table II are measured on the ETH/UCY benchmarks under leave-one-out evaluation, with the best task-loss variant (DTGAN-G) and random-weight scheme selected by comparison; selecting among model variants after seeing results is model selection, not a fitted input renamed as a prediction, and no metric is forced by construction. Self-citations appear only in background references on trajectory data and in the choice of the discriminator configuration, and no load-bearing argument reduces to a self-citation or to an imported uniqueness theorem. The limitation statement in Section IV-I explicitly concedes that the theoretical underpinning is not explored, which is an honest scope restriction rather than a circular justification. Overall, the central derivation chain is self-contained and externally benchmarked, so no circularity is present.

Assumptions & free parameters 6 free parameters · 5 assumptions · 1 invented entities

The method introduces random edge multipliers as the central mechanism, but their distribution, trainability, and independent effect are unspecified. Other free parameters include the task-loss weight, architecture dimensions, and sampling counts. The main domain assumptions are standard for GAN-based trajectory prediction, plus the ad hoc assumption that random multiplicative noise carries interaction meaning.

free parameters (6)
  • gamma task-loss weight = not reported
    Appears in Eq. (10), (11), and (13) balancing WGAN loss and task loss; value is never given.
  • random-weight distribution and range = not specified
    At is described as a randomly sampled matrix in Sec. III-A, but no distribution or bounds are provided; central to the method.
  • architecture dimensions (embedding dim F, GAT heads, TCN channels, CNN layers) = not reported
    Generator and discriminator sizes are not listed beyond kernel size 3 in Sec. IV-B.
  • epsilon in Eq. (12) = not reported
    Stabilizer in the Uniform density, added to avoid division by zero; value unspecified.
  • minPed = 3
    Minimum number of pedestrians per graph in Appendix Algorithm 1; chosen without sensitivity analysis.
  • K, number of samples = 20 for ADE/FDE, 100 for AMD/AMV
    Best-of-K evaluation; the choice affects reported errors and is not varied or justified.
assumptions (5)
  • domain assumption Pedestrian coordinates at each time step form graph nodes, and the graph adjacency represents social interactions.
    Sec. III-A and Fig. 1 construct G_t from coordinates and define A_t as an interaction matrix.
  • domain assumption WGAN objective combined with a task loss improves multi-modal trajectory realism and accuracy.
    Sec. III-D assumes adversarial training with task losses balances realism and ground-truth alignment.
  • ad hoc to paper A fixed random multiplicative mask on GAT attention can capture implicit social interactions.
    Eq. (3) multiplies attention by random e_ij; no derivation or independent evidence supports this assumption.
  • domain assumption Future trajectory coordinates follow a bivariate Gaussian distribution in DTGAN-G.
    Eq. (11) uses a Gaussian negative log-likelihood as the task loss.
  • domain assumption Data preprocessing retains only pedestrians present in all time slots of a window.
    Appendix Algorithm 1 skips pedestrians with missing frames, which may bias toward continuously visible pedestrians.
invented entities (1)
  • Random weight edge multiplier e_ij
    purpose: Replaces pre-defined interaction weights in graph adjacency to let the network learn implicit social interactions.
    No theoretical or independent dataset evidence that random multipliers encode interactions; only in-model comparison on ETH/UCY.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pedestrian Trajectory Prediction Based on Social Interactions Learning With Random Weights." pith.science (2026). https://pith.science/paper/VESOMU5S

@misc{pith2026250107711,
  author       = {Pith},
  title        = {Pith review of: Pedestrian Trajectory Prediction Based on Social Interactions Learning With Random Weights},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VESOMU5S}},
  note         = {Machine review of arXiv:2501.07711}
}
read the original abstract

Pedestrian trajectory prediction is a critical technology in the evolution of self-driving cars toward complete artificial intelligence. Over recent years, focusing on the trajectories of pedestrians to model their social interactions has surged with great interest in more accurate trajectory predictions. However, existing methods for modeling pedestrian social interactions rely on pre-defined rules, struggling to capture non-explicit social interactions. In this work, we propose a novel framework named DTGAN, which extends the application of Generative Adversarial Networks (GANs) to graph sequence data, with the primary objective of automatically capturing implicit social interactions and achieving precise predictions of pedestrian trajectory. DTGAN innovatively incorporates random weights within each graph to eliminate the need for pre-defined interaction rules. We further enhance the performance of DTGAN by exploring diverse task loss functions during adversarial training, which yields improvements of 16.7\% and 39.3\% on metrics ADE and FDE, respectively. The effectiveness and accuracy of our framework are verified on two public datasets. The experimental results show that our proposed DTGAN achieves superior performance and is well able to understand pedestrians' intentions.

Figures

Figures reproduced from arXiv: 2501.07711 by the authors.

Figure 1
Figure 1. Pedestrian interactions graph representation. We model pedes￾trian trajectories at each time t as graph-structured data, using different colored nodes to represent different pedestrians and edges to represent social interactions among pedestrians. future period. In the context of autonomous driving [7]–[10] and smart mobility [11], [12], vehicles are required to predict the trajectories of nearby pedestrians to avoi… view at source ↗
Figure 2
Figure 2. The DTGAN framework consists of a Generator and a Discriminator. Spatial Embedding Layer (SPE) is used to embed node features, and Random Weights (RW) to randomly weight the adjacency matrix for each graph. The Generator takes a set of graphs with node features and a random weights matrix as input. It utilizes the Graph Attention Network (GAT) to capture hidden node features and learn social interactions among pedes… view at source ↗
Figure 3
Figure 3. Illustration of single trajectory prediction. We use the coordinate system to represent the plane position of trajectory points, y is the vertical axis coordinate, x is the horizontal axis coordinate, and the trajectory point represents each moment. Two models use the best amongst 20 samples for evaluation. Note that the coordinate origin in each subplot is not exactly the same, and the intersection point does not n… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Illustration of trajectory distribution prediction. Each pedestrian is assigned various colors. The colored area of the ellipse represents the probability density distribution from the prediction. The wider the area of the ellipse, the greater the variance. the intenti…
Figure 5
Figure 5. Figure 5: Average ADE/FDE of all datasets for different generated ways of weights. F. Robustness Analysis about Random Weights We pre-define weights to validate the performance of DTGAN-G, as shown in [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Line chart of ADE/FDE using various random seeds. The x-axis [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Diverse scenarios from the ETH and UCY datasets illustrating different environments and crowd dynamics. [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

59 extracted references · 55 canonical work pages

  1. [1]

    Understanding private car aggregation effect via spatio-temporal analy- sis of trajectory data,

    Z. Xiao, H. Fang, H. Jiang, J. Bai, V . Havyarimana, H. Chen, and L. Jiao, “Understanding private car aggregation effect via spatio-temporal analy- sis of trajectory data,” IEEE Transactions on Cybernetics, vol. 53, no. 4, pp. 2346–2357, 2023. 1

    work page 2023

  2. [2]

    A Fusion Framework based on Sparse Gaussian-Wigner Prediction for Vehicle Localization using GDOP of GPS Satellites,

    V . Havyarimana, Z. Xiao, A. Sibomana, D. Wu, and J. Bai, “A Fusion Framework based on Sparse Gaussian-Wigner Prediction for Vehicle Localization using GDOP of GPS Satellites,” IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 2, pp. 680–689, 2020. 1

    work page 2020

  3. [3]

    An empirical study of travel behavior using private car trajectory data,

    H. Jiang, Y . Zhang, Z. Xiao, P. Zhao, and A. K. Iyengar, “An empirical study of travel behavior using private car trajectory data,” IEEE Trans- actions on Network Science and Engineering , vol. 8, no. 1, pp. 53–64,

    work page

  4. [4]

    Vehicle trajectory interpolation based on ensemble transfer regression,

    J. Xiao, Z. Xiao, D. Wang, V . Havyarimana, C. Liu, C. Zou, and D. Wu, “Vehicle trajectory interpolation based on ensemble transfer regression,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 7, pp. 7680–7691, 2022. 1

    work page 2022

  5. [5]

    Trajectory data acquisition via private car positioning based on tightly-coupled gps/obd integration in urban environments,

    Z. Xiao, Y . Chen, M. Alazab, and H. Chen, “Trajectory data acquisition via private car positioning based on tightly-coupled gps/obd integration in urban environments,” IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 7, pp. 9680–9691, 2022. 1

    work page 2022

  6. [6]

    Exploring individual travel patterns across private car trajectory data,

    Y . Huang, Z. Xiao, D. Wang, H. Jiang, and D. Wu, “Exploring individual travel patterns across private car trajectory data,” IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 12, pp. 5036–5050, 2020. 1

    work page 2020

  7. [7]

    Trajectory prediction for autonomous driving using spatial-temporal graph attention transformer,

    K. Zhang, X. Feng, L. Wu, and Z. He, “Trajectory prediction for autonomous driving using spatial-temporal graph attention transformer,” IEEE Transactions on Intelligent Transportation Systems , 2022. 1

    work page 2022

  8. [8]

    View Vertically: A Hierarchical Network for Trajectory Prediction via Fourier Spectrums

    C. Wong, B. Xia, Z. Hong, Q. Peng, W. Yuan, Q. Cao, Y . Yang, and X. You, “View vertically: A hierarchical network for trajectory prediction via fourier spectrums.” vol. abs/2110.07288, pp. 682–700,

    work page Pith review arXiv

Show all 59 references
  1. [9]

    Intent prediction of pedestrians via motion trajectories using stacked recurrent neural networks,

    K. Saleh, M. Hossny, and S. Nahavandi, “Intent prediction of pedestrians via motion trajectories using stacked recurrent neural networks,” IEEE Transactions on Intelligent Vehicles , vol. 3, no. 4, pp. 414–424, 2018. 1

    2018

  2. [10]

    Cscnet: Contex- tual semantic consistency network for trajectory prediction in crowded spaces,

    B. Xia, C. Wong, Q. Peng, W. Yuan, and X. You, “Cscnet: Contex- tual semantic consistency network for trajectory prediction in crowded spaces,” Pattern Recognition, vol. 126, p. 108552, 2022. 1

    2022

  3. [11]

    Intent prediction and trajectory forecasting via predictive inverse linear-quadratic regulation,

    M. Monfort, A. Liu, and B. Ziebart, “Intent prediction and trajectory forecasting via predictive inverse linear-quadratic regulation,” in Twenty- Ninth AAAI Conference on Artificial Intelligence , 2015, pp. 3672–3678. 1

    2015

  4. [12]

    Location prediction for individual vehicles via exploiting travel regularity and preference,

    W. Long, T. Li, Z. Xiao, D. Wang, R. Zhang, A. C. Regan, H. Chen, and Y . Zhu, “Location prediction for individual vehicles via exploiting travel regularity and preference,” IEEE Transactions on Vehicular Technology, vol. 71, no. 5, pp. 4718–4732, 2022. 1

    2022

  5. [13]

    Perception Task Offloading with Collaborative Computation for Autonomous Driving,

    Z. Xiao, J. Shu, H. Jiang, G. Min, H. Chen, and Z. Han, “Perception Task Offloading with Collaborative Computation for Autonomous Driving,” IEEE Journal on Selected Areas in Communications , vol. 41, no. 2, pp. 457–473, 2023. 1

    2023

  6. [14]

    Toward col- laborative occlusion-free perception in connected autonomous vehicles,

    Z. Xiao, J. Shu, H. Jiang, G. Min, J. Liang, and A. Iyengar, “Toward col- laborative occlusion-free perception in connected autonomous vehicles,” IEEE Transactions on Mobile Computing , pp. 1–13, 2023. 1

    2023

  7. [15]

    A deep spatiotemporal perspective for understanding crowd behavior,

    Y . Li, “A deep spatiotemporal perspective for understanding crowd behavior,” IEEE Transactions on multimedia, vol. 20, no. 12, pp. 3289– 3297, 2018. 1

    2018

  8. [16]

    Social lstm: Human trajectory prediction in crowded spaces,

    A. Alahi, K. Goel, V . Ramanathan, A. Robicquet, L. Fei-Fei, and S. Savarese, “Social lstm: Human trajectory prediction in crowded spaces,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 961–971. 1, 3, 5, 6

    2016

  9. [17]

    Social attention: Modeling attention in human crowds,

    A. Vemula, K. Muelling, and J. Oh, “Social attention: Modeling attention in human crowds,” in 2018 IEEE international Conference on Robotics and Automation (ICRA) . IEEE, 2018, pp. 4601–4607. 1

    2018

  10. [18]

    Encoding crowd interaction with deep neural network for pedestrian trajectory prediction,

    Y . Xu, Z. Piao, and S. Gao, “Encoding crowd interaction with deep neural network for pedestrian trajectory prediction,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 5275–5284. 1

    2018

  11. [19]

    Social gan: Socially acceptable trajectories with generative adversarial networks,

    A. Gupta, J. Johnson, L. Fei-Fei, S. Savarese, and A. Alahi, “Social gan: Socially acceptable trajectories with generative adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 2255–2264. 1, 3, 5, 6

    2018

  12. [20]

    Sophie: An attentive gan for predicting paths compliant to social and physical constraints,

    A. Sadeghian, V . Kosaraju, A. Sadeghian, N. Hirose, H. Rezatofighi, and S. Savarese, “Sophie: An attentive gan for predicting paths compliant to social and physical constraints,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 1349–

    2019

  13. [21]

    Social-bigat: Multimodal trajectory forecasting using 12 bicycle-gan and graph attention networks,

    V . Kosaraju, A. Sadeghian, R. Martín-Martín, I. Reid, H. Rezatofighi, and S. Savarese, “Social-bigat: Multimodal trajectory forecasting using 12 bicycle-gan and graph attention networks,” Advances in Neural Infor- mation Processing Systems , vol. 32, pp. 137–146, 2019. 1, 2, 3, 6

    2019

  14. [22]

    Stgat: Modeling spatial- temporal interactions for human trajectory prediction,

    Y . Huang, H. Bi, Z. Li, T. Mao, and Z. Wang, “Stgat: Modeling spatial- temporal interactions for human trajectory prediction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2019, pp. 6272–6281. 2, 3, 6

    2019

  15. [23]

    Social-stgcnn: A social spatio-temporal graph convolutional neural network for human trajectory prediction,

    A. Mohamed, K. Qian, M. Elhoseiny, and C. Claudel, “Social-stgcnn: A social spatio-temporal graph convolutional neural network for human trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 14 424–14 432. 2, 5, 6

    2020

  16. [24]

    Conditional generative adversarial nets,

    M. Mirza and S. Osindero, “Conditional generative adversarial nets,” Computer Science, pp. 2672–2680, 2014. 2

    2014

  17. [25]

    Foreseeing private car transfer between urban regions with multiple graph-based generative adversarial networks,

    C. Liu, Z. Xiao, D. Wang, M. Cheng, H. Chen, and J. Cai, “Foreseeing private car transfer between urban regions with multiple graph-based generative adversarial networks,” World Wide Web, vol. 25, no. 6, pp. 2515–2534, 2022. 2

    2022

  18. [26]

    Recognition-oriented image compressive sensing with deep learning,

    S. Zhou, X. Deng, C. Li, Y . Liu, and H. Jiang, “Recognition-oriented image compressive sensing with deep learning,” IEEE Transactions on Multimedia, 2022. 2

    2022

  19. [27]

    A new perspective on stabilizing gans training: Direct adversarial training,

    Z. Li, P. Xia, R. Tao, H. Niu, and B. Li, “A new perspective on stabilizing gans training: Direct adversarial training,” IEEE Transactions on Emerging Topics in Computational Intelligence , pp. 1–12, 2022. 2

    2022

  20. [28]

    A comprehensive survey on data-efficient gans in image generation,

    Z. Li, X. Wu, B. Xia, J. Zhang, C. Wang, and B. Li, “A comprehensive survey on data-efficient gans in image generation,” CoRR, vol. abs/2204.08329, 2022. [Online]. Available: https://doi.org/10.48550/ arXiv.2204.08329 2

    2022 doi arXiv

  21. [29]

    Fakeclr: Exploring contrastive learning for solving latent discontinuity in data- efficient gans,

    Z. Li, C. Wang, H. Zheng, J. Zhang, and B. Li, “Fakeclr: Exploring contrastive learning for solving latent discontinuity in data- efficient gans,” CoRR, vol. abs/2207.08630, 2022. [Online]. Available: https://doi.org/10.48550/arXiv.2207.08630 2

    2022 doi arXiv

  22. [30]

    A systematic survey of regularization and normalization in gans,

    Z. Li, M. Usman, R. Tao, P. Xia, C. Wang, H. Chen, and B. Li, “A systematic survey of regularization and normalization in gans,” ACM Comput. Surv. , vol. 55, no. 11, pp. 232:1–232:37, 2023. [Online]. Available: https://doi.org/10.1145/3569928 2

    2023 doi

  23. [31]

    Social force model for pedestrian dynamics,

    D. Helbing and P. Molnar, “Social force model for pedestrian dynamics,” Physical review E , vol. 51, no. 5, p. 4282, 1995. 2

    1995

  24. [32]

    Abnormal crowd behavior detec- tion using social force model,

    R. Mehran, A. Oyama, and M. Shah, “Abnormal crowd behavior detec- tion using social force model,” in 2009 IEEE conference on computer vision and pattern recognition . IEEE, 2009, pp. 935–942. 2

    2009

  25. [33]

    Everybody needs somebody: Modeling social and grouping behavior on a linear program- ming multiple people tracker,

    L. Leal-Taixé, G. Pons-Moll, and B. Rosenhahn, “Everybody needs somebody: Modeling social and grouping behavior on a linear program- ming multiple people tracker,” in 2011 IEEE international conference on computer vision workshops (ICCV workshops) . IEEE, 2011, pp. 120–127. 2

    2011

  26. [34]

    People tracking with human motion predictions from social forces,

    M. Luber, J. A. Stork, G. D. Tipaldi, and K. O. Arras, “People tracking with human motion predictions from social forces,” in 2010 IEEE international conference on robotics and automation . IEEE, 2010, pp. 464–469. 2

    2010

  27. [35]

    Improving data association by joint modeling of pedestrian trajectories and groupings,

    S. Pellegrini, A. Ess, and L. V . Gool, “Improving data association by joint modeling of pedestrian trajectories and groupings,” in European conference on computer vision . Springer, 2010, pp. 452–465. 2

    2010

  28. [36]

    Social force model-based mcmc-ocsvm particle phd filter for multiple human tracking,

    P. Feng, W. Wang, S. Dlay, S. M. Naqvi, and J. Chambers, “Social force model-based mcmc-ocsvm particle phd filter for multiple human tracking,” IEEE Transactions on Multimedia , vol. 19, no. 4, pp. 725– 739, 2016. 2

    2016

  29. [37]

    Discrete choice models of pedestrian walking behavior,

    G. Antonini, M. Bierlaire, and M. Weber, “Discrete choice models of pedestrian walking behavior,” Transportation Research Part B: Method- ological, vol. 40, no. 8, pp. 667–687, 2006. 2

    2006

  30. [38]

    Gaussian process dynamical models for human motion,

    J. M. Wang, D. J. Fleet, and A. Hertzmann, “Gaussian process dynamical models for human motion,” IEEE transactions on pattern analysis and machine intelligence, vol. 30, no. 2, pp. 283–298, 2007. 2

    2007

  31. [39]

    Modelling smooth paths using gaussian processes,

    M. K. C. Tay and C. Laugier, “Modelling smooth paths using gaussian processes,” in Field and Service Robotics. Springer, 2008, pp. 381–390. 2

    2008

  32. [40]

    Semi-supervised classification with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classification with graph convolutional networks,” in 5th International Conference on Learning Representations (ICLR), 2017. 2

    2017

  33. [41]

    Spatio-temporal graph transformer networks for pedestrian trajectory prediction,

    C. Yu, X. Ma, J. Ren, H. Zhao, and S. Yi, “Spatio-temporal graph transformer networks for pedestrian trajectory prediction,” in European Conference on Computer Vision . Springer, 2020, pp. 507–523. 2, 6

    2020

  34. [42]

    Graph-based spatial transformer with memory replay for multi-future pedestrian trajectory prediction,

    L. Li, M. Pagnucco, and Y . Song, “Graph-based spatial transformer with memory replay for multi-future pedestrian trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 2231–2241. 2

    2022

  35. [43]

    Adversarial-example at- tacks toward android malware detection system,

    H. Li, S. Zhou, W. Yuan, J. Li, and H. Leung, “Adversarial-example at- tacks toward android malware detection system,” IEEE Systems Journal, vol. 14, no. 1, pp. 653–656, 2019. 3

    2019

  36. [44]

    Black- box adversarial example attack towards fcg based android malware detection under incomplete feature information,

    H. Li, Z. Cheng, B. Wu, L. Yuan, C. Gao, W. Yuan, and X. Luo, “Black- box adversarial example attack towards fcg based android malware detection under incomplete feature information,” vol. abs/2303.08509, 2023, pp. 1181–1198. 3

    2023 arXiv

  37. [45]

    Graph attention networks,

    P. Veli ˇckovi´c, G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y . Bengio, “Graph attention networks,” in 6th International Conference on Learning Representations (ICLR) , 2018. 3

    2018

  38. [46]

    Unified spatial-temporal neighbor attention network for dynamic traffic prediction,

    W. Long, Z. Xiao, D. Wang, H. Jiang, J. Chen, Y . Li, and M. Alazab, “Unified spatial-temporal neighbor attention network for dynamic traffic prediction,” IEEE Transactions on Vehicular Technology, vol. 72, no. 2, pp. 1515–1529, 2022. 3

    2022

  39. [47]

    An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,

    S. Bai, J. Z. Kolter, and V . Koltun, “An empirical evaluation of generic convolutional and recurrent networks for sequence modeling,” CoRR, vol. abs/1803.01271, 2018. [Online]. Available: http://arxiv.org/abs/1803.01271 4

    2018 arXiv

  40. [48]

    Ace-gcn: A fast data-driven fpga accelerator for gcn embedding,

    J. Romero Hung, C. Li, P. Wang, C. Shao, J. Guo, J. Wang, and G. Shi, “Ace-gcn: A fast data-driven fpga accelerator for gcn embedding,” ACM Transactions on Reconfigurable Technology and Systems (TRETS), vol. 14, no. 4, pp. 1–23, 2021. 4

    2021

  41. [49]

    Dragon: dynamic recurrent accelerator for graph online convolution,

    J. Romero Hung, C. Li, T. Wang, J. Guo, P. Wang, C. Shao, J. Wang, G. Shi, X. Liu, and H. Wu, “Dragon: dynamic recurrent accelerator for graph online convolution,” ACM Transactions on Design Automation of Electronic Systems, vol. 28, no. 1, pp. 1–27, 2023. 4

    2023

  42. [50]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” Advances in neural information processing systems , vol. 27, 2014. 5

    2014

  43. [51]

    Wasserstein generative adversarial networks,

    M. Arjovsky, S. Chintala, and L. Bottou, “Wasserstein generative adversarial networks,” in Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017 , ser. Proceedings of Machine Learning Research, D. Precup and Y...

    2017

  44. [52]

    Are gans created equal? a large-scale study,

    M. Lucic, K. Kurach, M. Michalski, S. Gelly, and O. Bousquet, “Are gans created equal? a large-scale study,” in Advances in Neural Information Processing Systems , S. Bengio, H. Wallach, H. Larochelle, K. Grauman, N. Cesa-Bianchi, and R. Garnett, Eds., vol. 31. Curran Associat...

    2018

  45. [53]

    Temporal pyramid network for pedestrian trajectory prediction with multi-supervision,

    R. Liang, Y . Li, X. Li, Y . Tang, J. Zhou, and W. Zou, “Temporal pyramid network for pedestrian trajectory prediction with multi-supervision,” in Proceedings of the AAAI conference on artificial intelligence , vol. 35, no. 3, 2021, pp. 2029–2037. 6

    2021

  46. [54]

    Br- gan: A pedestrian trajectory prediction model combined with behavior recognition,

    S. M. Pang, J. X. Cao, M. Y . Jian, J. Lai, and Z. Y . Yan, “Br- gan: A pedestrian trajectory prediction model combined with behavior recognition,” IEEE Transactions on Intelligent Transportation Systems ,

  47. [56]

    You’ll never walk alone: Modeling social behavior for multi-target tracking,

    S. Pellegrini, A. Ess, K. Schindler, and L. Van Gool, “You’ll never walk alone: Modeling social behavior for multi-target tracking,” in 2009 IEEE 12th international conference on computer vision . IEEE, 2009, pp. 261–268. 5

    2009

  48. [57]

    Learning an image-based motion context for multiple people tracking,

    L. Leal-Taixé, M. Fenzi, A. Kuznetsova, B. Rosenhahn, and S. Savarese, “Learning an image-based motion context for multiple people tracking,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2014, pp. 3542–3549. 5

    2014

  49. [58]

    Social-implicit: Rethinking trajectory prediction evaluation and the effectiveness of implicit maximum likelihood estimation,

    A. A. Mohamed, D. Zhu, W. Vu, M. Elhoseiny, and C. G. Claudel, “Social-implicit: Rethinking trajectory prediction evaluation and the effectiveness of implicit maximum likelihood estimation,” in Computer Vision - ECCV 2022 - 17th European Conference, Tel Aviv, Israel, October 2...

    2022

  50. [59]

    Gcn-gan: A non-linear temporal link prediction model for weighted dynamic networks,

    K. Lei, M. Qin, B. Bai, G. Zhang, and M. Yang, “Gcn-gan: A non-linear temporal link prediction model for weighted dynamic networks,” IEEE INFOCOM 2019 - IEEE Conference on Computer Communications , pp. 388–396, 2019. 5 13 Jiajia Xie received the M.S. degree from the College of...

    2019

  51. [2022]

    Available: https://arxiv.org/abs/2202.03954 6

    [Online]. Available: https://arxiv.org/abs/2202.03954 6

    arXiv

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.