Pith. sign in

REVIEW 5 major objections 5 minor 55 references

SocialMOIF: Multi-Order Intention Fusion for Pedestrian Trajectory Prediction

T0 review · 5 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read SocialMOIF claims that fusing direct and indirect neighbor intentions yields state-of-the-art pedestrian trajectory predictions across ETH/UCY, NBA, SDD, and NuScenes.

desk verdict Strong reported results on multiple benchmarks, but a load-bearing sign error in the direction loss makes the paper not reproducible as written; needs correction and variance reporting. read the letter →

arxiv 2504.15616 v1 pith:5YXYTOET submitted 2025-04-22 cs.LG cs.CV

Kai Chen , Xiaodong Zhao , Yujie Huang , Guoyu Fang , Xiao Song , Ruiping Wang , Ziyuan Wang This is my paper · ORCID
classification cs.LGcs.CV
keywords pedestriantrajectorypredictionmulti-orderintentionfusionhigher-ordersocialinteractiondistributionapproximatorKolmogorov-ArnoldNetworksdistance-directionlossADE/FDEgenerativemodels
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 tries to establish that pedestrian trajectory prediction improves when a model distinguishes two kinds of social intention: the direct influence of each neighbor on the target pedestrian and the indirect influence that neighbors exert on one another before reaching the target. To capture both, the authors build SocialMOIF, which fuses a first-order intention layer with a higher-order intention layer and then uses the fused intention to guide a latent trajectory distribution, refines predictions with a Kolmogorov-Arnold Network optimizer, and trains with a distance-direction loss. If the reported benchmark numbers hold, the model would be the new state of the art on ETH/UCY, the NBA Rebound and Scores subsets, SDD, and NuScenes in best-of-20 average and final displacement error, bettering the previous best ADE/FDE on every one of those benchmarks. The paper's central contribution is a concrete architecture for including indirect group influence alongside direct neighbor interaction without letting higher-order effects overwhelm first-order ones.

What carries the argument

The load-bearing object is the Multi-Order Intention Fusion (MOIF) module. Its first-order layer computes an attention matrix between the target and each neighbor; its higher-order layer computes a multi-head self-attention matrix among the neighbors themselves, which the paper justifies by decomposing group intentions into pairwise first-order interactions. The two matrices are combined by adding the higher-order matrices scaled by learned influence factors to the first-order matrix, then using the result to attend to value vectors. The fused intention feeds a VAE-style trajectory distribution approximator, where an RNN updates the latent variable and a future-trajectory embedding serves as an upper bound; the decoder output is then refined by a KAN-based global trajectory optimizer operating over the entire time dimension at once, and the full system is supervised by the sum of displacement error and angle error.

What would settle it

Run the identical SocialMOIF pipeline on ETH/UCY, NBA, SDD, and NuScenes with the higher-order neighbor-neighbor attention layer removed or replaced by a same-capacity non-interacting mixer, and compare best-of-20 ADE/FDE; if the numbers stay within noise, the claimed higher-order intention mechanism is not what drives the reported gains.

Watch

Extended reading notes

Core claim

The paper's central discovery is that a multi-order intention fusion layer—one head for direct target-neighbor interactions, one head for neighbor-neighbor interactions within the group, combined with per-subspace learned weights—can reduce prediction errors across datasets that differ in crowd density, motion purpose, and static or dynamic scenes. In the paper's tables, best-of-20 ADE/FDE drop from 0.15/0.20 to 0.13/0.18 on ETH/UCY, from 0.54/0.79 to 0.34/0.66 on NBA Rebound, from 0.46/0.76 to 0.30/0.56 on NBA Scores, from 0.21/0.34 to 0.17/0.24 on SDD, and from 1.04/1.64 to 0.92/1.56 on NuScenes. The paper attributes these gains to four interacting components: the Multi-Order Intention Fusion module, a trajectory distribution approximator that explicitly models the latent intent distribution during training, a global optimizer built from Kolmogorov-Arnold Networks that predicts all future time steps in parallel, and a loss function that penalizes both distance error and direction mismatch.

Load-bearing premise

The model's higher-order layer is built from pairwise self-attention among neighbors, so the central claim depends on group-level effects on a pedestrian being fully decomposable into pairwise neighbor-to-neighbor effects.

Editorial extensions

If this is right

  • If the reported results reproduce, SocialMOIF becomes the best-published best-of-20 ADE/FDE model on ETH/UCY, NBA Rebound, NBA Scores, SDD, and NuScenes at the time of the work.
  • The explicit latent-variable updating scheme means the training-time trajectory distribution is directly supervised rather than left implicit, which the paper argues improves interpretability.
  • Predicting all future frames in parallel through the KAN-based optimizer removes sequential decoding dependency, reducing error accumulation and enabling faster inference.
  • The directional loss term is claimed to improve other models as well when added, making it a separable training recipe independent of the rest of SocialMOIF.

Reading between the lines

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

  • Beyond the paper, because the higher-order layer is mathematically a set of pairwise attention operations, the reported gains could come from extra model capacity rather than from true group-level reasoning; swapping that layer for a same-capacity non-interacting mixer would settle this.
  • Beyond the paper, the 'squeeze theorem' description is an analogy unless one proves that the sampled latent distribution is actually bounded by the fused intention and the future-trajectory embedding; checking that bound empirically would strengthen the interpretability claim.
  • Beyond the paper, the largest relative ADE gain in the table is on NBA Rebound, a strongly goal-directed dataset, which suggests the approach may show its clearest value in sports and other purposeful-motion settings, and long-horizon team-sport prediction would be a natural stress test.
  • Beyond the paper, best-of-20 ADE/FDE can reward a wide proposal distribution rather than per-mode accuracy; reporting best-of-1 and per-sample NLL alongside best-of-20 would show whether the improvement is genuine accuracy or mostly diversity.
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 SocialMOIF, a generative trajectory prediction model whose stated contributions are a multi-order intention fusion module (first-order target-neighbor interactions plus higher-order neighbor-neighbor interactions), a trajectory distribution approximator based on latent variables, a KAN-based global trajectory optimizer for parallel decoding, and a distance-direction fused loss. The model is evaluated on ETH/UCY, NBA, SDD, and NuScenes with best-of-20 ADE/FDE metrics, and the paper reports state-of-the-art results on essentially every benchmark, with ablations attributing gains to each component. The central claim is that SocialMOIF outperforms previous state-of-the-art baselines across multiple metrics in both dynamic and static datasets.

Significance. If the reported numbers are reproducible, this would be a strong empirical result: the improvements on NBA (e.g., Rebound ADE 0.34 vs 0.54), SDD (ADE 0.17 vs 0.21), and NuScenes (ADE 0.92 vs 1.04) are substantial. The paper also ships a code link, compares with a wide set of baselines, and provides component-level ablations, which are strengths. However, the significance is currently conditional on resolving serious internal inconsistencies in the definition of the loss and the fusion formula, and on providing statistical support for the headline claims.

major comments (5)
  1. [3.4, Eqs. (16)-(18)] The optimization objective is internally inconsistent as written. Langle in Eq. (17) is defined as minus the arccosine of a cosine similarity, so Langle is non-positive; if L in Eq. (18) is minimized, the direction term repels the predicted displacement from the ground-truth direction, and if L is instead maximized as an ELBO, the positive distance term Ldis would penalize accurate predictions. Since the ablation in Table 2 (groups 6 vs 7) attributes a large improvement to the direction component A, the paper must state which sign is actually used in training and correct the equations accordingly.
  2. [3.1, Eq. (8)] The fusion rule in Eq. (8) adds matrices of incompatible shapes: W_U^m is an Nn x Nn neighbor self-attention matrix from Eq. (3), while W_S is the target-to-neighbor attention matrix from Eq. (7), which has one target query against Nn neighbor keys and therefore shape 1 x Nn. Unless an omitted reshaping or broadcasting is specified, the sum is undefined, so the central multi-order fusion formula cannot be implemented as written.
  3. [4, Tables 1-2] No error bars, confidence intervals, or multiple-seed results are reported for any experiment. Some headline differences are small (e.g., NuScenes FDE 1.56 vs 1.64, and on ETH the proposed method is actually worse than V2-Net-SC on ADE), so without repeated runs and variance reporting the claimed state-of-the-art status is not statistically supported.
  4. [3.1, Higher-order intention interaction layer] The paper's conceptual claim that higher-order intention interactions are captured is not supported by the implementation. The layer computes pairwise self-attention among neighbors via Eq. (3), and the stated decomposition assumption ('higher-order intention interactions within the group of neighbors can be decomposed into first-order intention interactions between individual neighbors') is asserted without justification or an ablation that varies the number of interaction hops or compares against a hypergraph/relational baseline. If the claim is meant as an approximation, it should be validated; otherwise the contribution should be framed as pairwise neighbor interaction modeling rather than higher-order fusion.
  5. [4, Table 1] The directly related higher-order relational reasoning method of Kim et al. [16] is cited in the introduction but is not included in any comparison table. Given the paper's central higher-order-intention claim, omitting the most relevant baseline makes it difficult to assess the marginal value of the proposed multi-order fusion.
minor comments (5)
  1. [3.2, Eq. (10)] Equation (10) and the surrounding text are difficult to parse; the reparameterized latent variable equation mixes the definition and the conditional distribution in one line. Please rewrite with a clear definition of q_phi and the sampling operation.
  2. [3.4, Eq. (18)] Equation (18) contains a typographical artifact 'TFP' before the summation; it should be a summation symbol over the prediction horizon.
  3. [4.1, NuScenes results] The reported NuScenes FDE improvement is inconsistent with Table 1: the text states a 1.87% reduction, but Table 1 lists 1.56 vs 1.64 (4.9%) if compared to E-V2-Net-SC or 1.56 vs 1.59 (1.9%) if compared to V2-Net-SC. Please reconcile the text and the table.
  4. [4, Metrics] The NLL metric is mentioned in Section 4 as a reported metric, but all NLL results are deferred to a Supplementary Material that is not included with the submission; either include the results or explicitly state that they are omitted.
  5. [3.4, Eq. (17)] In Eq. (17), the direction vectors are written as \hat g_i^t - \hat g_i^{t+1} rather than the standard \hat g_i^{t+1} - \hat g_i^t; although the predicted and ground-truth vectors are both reversed so the angle is unchanged, the notation is confusing and should be corrected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is a standard supervised generative pipeline with external benchmark comparisons; the only notable issue is a loss-sign inconsistency, which is a correctness concern, not circularity.

full rationale

The paper's central claims are empirical SOTA results against external baselines, and the model is a conventional supervised generative architecture. The multi-order intention fusion is built from standard attention layers whose 'higher-order' decomposition is an explicit modeling assumption, not a result derived from the target metric. The trajectory distribution approximator is a conditional VAE: during training it conditions on the actual future trajectory B_i as a posterior/teacher-forcing input, and the paper explicitly states 'In test mode, B_i^t is removed', so no test-time prediction is secretly conditioned on ground truth. The global trajectory optimizer is a KAN module and the distance-direction loss supervises training with ground-truth values; none of these components is a parameter fitted to a subset of data and then renamed a prediction. The few self-citations are prior baselines or standard leave-one-out methodology, and they are not load-bearing uniqueness theorems or ansatz-smuggling citations. The sign inconsistency in Eqs. 16-18 (whether L is minimized or maximized, Langle pushes away from the correct direction or Ldis rewards large error) is a serious correctness/reproducibility flaw and should be flagged as such, but it is not a circular derivation: the prediction does not reduce by construction to its inputs.

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

The model introduces no new physical or conceptual entities. The free parameters are standard network hyperparameters and trainable fusion weights. The main assumption is the pairwise decomposition of higher-order interactions, which is asserted without proof or dedicated ablation.

free parameters (3)
  • Influence factors eta_m for higher-order subspaces = learned, m=1..6
    In Eq. (8), each of the M=6 subspaces has a trainable scalar eta_m that weights the contribution of the higher-order interaction matrix W_U^m to the fused attention score. These are learned from data and are central to the fusion mechanism.
  • Number of subspaces M = 6
    Set by hand in Section 4 (Implementation Details). The value is not derived and directly determines the capacity of the higher-order layer.
  • Depth of KAN optimizer L = 3
    Set by hand in Section 4. The depth controls how many KAN layers are applied in the global trajectory optimizer.
assumptions (3)
  • domain assumption Higher-order intention interactions among neighbors decompose into pairwise first-order interactions.
    Section 3.1 states this decomposition explicitly and implements the higher-order layer as self-attention, which is pairwise. The validity of this assumption is not tested, and it is the conceptual load-bearing premise for labeling the model 'multi-order'.
  • standard math Variational inference with the reparameterization trick provides a valid lower bound for the trajectory distribution.
    Eq. (10) and the loss in Eq. (18) use the standard VAE reparameterization and ELBO. This is standard, but the paper cites an unrelated reference [17] for the reparameterization trick instead of Kingma and Welling.
  • standard math KAN layers can approximate the global trajectory transformation to a sufficient accuracy.
    The global optimizer in Eq. (14)-(15) treats the KAN as a universal function approximator, following Liu et al. [23, 24]. No specific approximation bound is derived.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SocialMOIF: Multi-Order Intention Fusion for Pedestrian Trajectory Prediction." pith.science (2026). https://pith.science/paper/5YXYTOET

@misc{pith2026250415616,
  author       = {Pith},
  title        = {Pith review of: SocialMOIF: Multi-Order Intention Fusion for Pedestrian Trajectory Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5YXYTOET}},
  note         = {Machine review of arXiv:2504.15616}
}
read the original abstract

The analysis and prediction of agent trajectories are crucial for decision-making processes in intelligent systems, with precise short-term trajectory forecasting being highly significant across a range of applications. Agents and their social interactions have been quantified and modeled by researchers from various perspectives; however, substantial limitations exist in the current work due to the inherent high uncertainty of agent intentions and the complex higher-order influences among neighboring groups. SocialMOIF is proposed to tackle these challenges, concentrating on the higher-order intention interactions among neighboring groups while reinforcing the primary role of first-order intention interactions between neighbors and the target agent. This method develops a multi-order intention fusion model to achieve a more comprehensive understanding of both direct and indirect intention information. Within SocialMOIF, a trajectory distribution approximator is designed to guide the trajectories toward values that align more closely with the actual data, thereby enhancing model interpretability. Furthermore, a global trajectory optimizer is introduced to enable more accurate and efficient parallel predictions. By incorporating a novel loss function that accounts for distance and direction during training, experimental results demonstrate that the model outperforms previous state-of-the-art baselines across multiple metrics in both dynamic and static datasets.

Figures

Figures reproduced from arXiv: 2504.15616 by the authors.

Figure 1
Figure 1. The process of SocialMOIF includes: (1) extracting the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Computational pipeline of SocialMOIF. Firstly, the first-order interaction layer focuses on direct interactions between the target [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Representative cases. Row 1 represented complex his￾torical trajectory cases. Row 2 represented state maintenance cases. Row 3 represented numerous neighbors cases. Unlike previous studies [6, 26, 39–41, 45, 46], the focus here was on conducting experiments involving historical [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Predictions comparison of predictions in ETH/UCY. The comparison included heatmaps of the overall distribution of predictions [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Distribution heatmaps. Every heatmap of future trajectory distribution was shown above the real scene image. (a) represented the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Representative cases about the optimization effects of [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 45 canonical work pages

  1. [16]

    Higher-order relational reasoning for pedestrian trajectory prediction

    Sungjune Kim, Hyung-gun Chi, Hyerin Lim, Karthik Ra- mani, Jinkyu Kim, and Sangpil Kim. Higher-order relational reasoning for pedestrian trajectory prediction. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15251–15260, 2024. 2

    2024

  2. [1]

    So- cial lstm: Human trajectory prediction in crowded spaces

    Alexandre Alahi, Kratarth Goel, Vignesh Ramanathan, Alexandre Robicquet, Li Fei-Fei, and Silvio Savarese. So- cial lstm: Human trajectory prediction in crowded spaces. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 961–971, 2016. 2, 3

    work page 2016

  3. [2]

    An evaluation of trajectory prediction approaches and notes on the trajnet benchmark

    Stefan Becker, Ronny Hug, Wolfgang H ¨ubner, and Michael Arens. An evaluation of trajectory prediction approaches and notes on the trajnet benchmark. arXiv preprint arXiv:1805.07663, 2018. 6

    arXiv 2018

  4. [3]

    Bayesian intention inference for trajectory prediction with an unknown goal destination

    Graeme Best and Robert Fitch. Bayesian intention inference for trajectory prediction with an unknown goal destination. In 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 5817–5823. IEEE, 2015. 2

    work page 2015

  5. [4]

    nuscenes: A multi- modal dataset for autonomous driving

    Holger Caesar, Varun Bankiti, Alex H Lang, Sourabh V ora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Gi- ancarlo Baldan, and Oscar Beijbom. nuscenes: A multi- modal dataset for autonomous driving. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11621–11631, 2020. 5

    2020

  6. [5]

    Pedestrian trajec- tory prediction in heterogeneous traffic using pose keypoints- based convolutional encoder-decoder network

    Kai Chen, Xiao Song, and Xiaoxiang Ren. Pedestrian trajec- tory prediction in heterogeneous traffic using pose keypoints- based convolutional encoder-decoder network. IEEE Trans- actions on Circuits and Systems for Video Technology, 31(5): 1764–1775, 2020. 2, 6

    work page 2020

  7. [6]

    Fully convolutional encoder-decoder with an attention mech- anism for practical pedestrian trajectory prediction

    Kai Chen, Xiao Song, Haitao Yuan, and Xiaoxiang Ren. Fully convolutional encoder-decoder with an attention mech- anism for practical pedestrian trajectory prediction. IEEE Transactions on Intelligent Transportation Systems, 23(11): 20046–20060, 2022. 2, 6

    work page 2022

  8. [7]

    Fu- ture pedestrian location prediction in first-person videos for autonomous vehicles and social robots

    Kai Chen, Haihua Zhu, Dunbing Tang, and Kun Zheng. Fu- ture pedestrian location prediction in first-person videos for autonomous vehicles and social robots. Image and Vision Computing, 134:104671, 2023. 2

    work page 2023

Show all 55 references
  1. [8]

    Msr-gcn: Multi-scale residual graph convolution networks for human motion prediction

    Lingwei Dang, Yongwei Nie, Chengjiang Long, Qing Zhang, and Guiqing Li. Msr-gcn: Multi-scale residual graph convolution networks for human motion prediction. In Pro- ceedings of the IEEE/CVF international conference on com- puter vision, pages 11467–11476, 2021. 3

    2021

  2. [9]

    Transformer networks for trajectory forecasting

    Francesco Giuliari, Irtiza Hasan, Marco Cristani, and Fabio Galasso. Transformer networks for trajectory forecasting. In 2020 25th international conference on pattern recognition (ICPR), pages 10335–10342. IEEE, 2021. 2

    2020

  3. [10]

    Stochastic trajectory prediction via motion indeterminacy diffusion

    Tianpei Gu, Guangyi Chen, Junlong Li, Chunze Lin, Yong- ming Rao, Jie Zhou, and Jiwen Lu. Stochastic trajectory prediction via motion indeterminacy diffusion. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17113–17122, 2022. 3, 6

    2022

  4. [11]

    Social gan: Socially acceptable tra- jectories with generative adversarial networks

    Agrim Gupta, Justin Johnson, Li Fei-Fei, Silvio Savarese, and Alexandre Alahi. Social gan: Socially acceptable tra- jectories with generative adversarial networks. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 2255–2264, 2018. 2, 3

    2018

  5. [12]

    Social force model for pedestrian dynamics

    Dirk Helbing and Peter Molnar. Social force model for pedestrian dynamics. Physical review E, 51(5):4282, 1995. 2

    1995

  6. [13]

    Simulation of pedes- trian flows by optimal control and differential games

    Serge Hoogendoorn and Piet HL Bovy. Simulation of pedes- trian flows by optimal control and differential games. Opti- mal control applications and methods, 24(3):153–172, 2003. 2

    2003

  7. [14]

    Modeling human behavior in vessel maneuver simulation by optimal control and game theory

    SP Hoogendoorn, Winnie Daamen, Yaqing Shu, and Han Ligteringen. Modeling human behavior in vessel maneuver simulation by optimal control and game theory. Transporta- tion research record, 2326(1):45–53, 2013. 2

    2013

  8. [15]

    Planning-oriented autonomous driving

    Yihan Hu, Jiazhi Yang, Li Chen, Keyu Li, Chonghao Sima, Xizhou Zhu, Siqi Chai, Senyao Du, Tianwei Lin, Wenhai Wang, et al. Planning-oriented autonomous driving. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17853–17862, 2023. 1

    2023

  9. [17]

    Machine learn- ing for emergency management: A survey and future outlook

    Christos Kyrkou, Panayiotis Kolios, Theocharis Theocharides, and Marios Polycarpou. Machine learn- ing for emergency management: A survey and future outlook. Proceedings of the IEEE, 111(1):19–41, 2022. 5

    2022

  10. [18]

    Crowds by example

    Alon Lerner, Yiorgos Chrysanthou, and Dani Lischinski. Crowds by example. In Computer graphics forum , pages 655–664. Wiley Online Library, 2007. 5

    2007

  11. [19]

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

    Lihuan Li, Maurice Pagnucco, and Yang 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, pages 2221–2231, 2022. 2, 6

    2022

  12. [20]

    Beyond minimum-of-n: Rethinking the eval- uation and methods of pedestrian trajectory prediction.IEEE Transactions on Circuits and Systems for Video Technology,

    Linhui Li, Xiaotong Lin, Yejia Huang, Zizhen Zhang, and Jian-Fang Hu. Beyond minimum-of-n: Rethinking the eval- uation and methods of pedestrian trajectory prediction.IEEE Transactions on Circuits and Systems for Video Technology,

  13. [21]

    Nba-player-movements

    Kostya Linou. Nba-player-movements. Software for Visual- ization of NBA Games, 2016. 5

    2016

  14. [22]

    Avgcn: Trajectory prediction using graph convolutional net- works guided by human attention

    Congcong Liu, Yuying Chen, Ming Liu, and Bertram E Shi. Avgcn: Trajectory prediction using graph convolutional net- works guided by human attention. In 2021 IEEE Inter- national Conference on Robotics and Automation (ICRA) , pages 14234–14240. IEEE, 2021. 3

    2021

  15. [23]

    Kan 2.0: Kolmogorov-arnold net- works meet science

    Ziming Liu, Pingchuan Ma, Yixuan Wang, Wojciech Ma- tusik, and Max Tegmark. Kan 2.0: Kolmogorov-arnold net- works meet science. arXiv preprint arXiv:2408.10205, 2024. 2

    2024 arXiv

  16. [24]

    Kan: Kolmogorov-arnold networks

    Ziming Liu, Yixuan Wang, Sachin Vaidya, Fabian Ruehle, James Halverson, Marin Solja ˇci´c, Thomas Y Hou, and Max Tegmark. Kan: Kolmogorov-arnold networks. arXiv preprint arXiv:2404.19756, 2024. 2

    2024 arXiv

  17. [25]

    It is not the journey but the destination: End- point conditioned trajectory prediction

    Karttikeya Mangalam, Harshayu Girase, Shreyas Agarwal, Kuan-Hui Lee, Ehsan Adeli, Jitendra Malik, and Adrien Gaidon. It is not the journey but the destination: End- point conditioned trajectory prediction. In Computer Vision– ECCV 2020: 16th European Conference, Glasgow, UK, A...

    2020

  18. [26]

    From goals, waypoints & paths to long term hu- man trajectory forecasting

    Karttikeya Mangalam, Yang An, Harshayu Girase, and Jiten- dra Malik. From goals, waypoints & paths to long term hu- man trajectory forecasting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 15233– 15242, 2021. 1, 2, 6

    2021

  19. [27]

    Leapfrog diffusion model for stochastic trajectory prediction

    Weibo Mao, Chenxin Xu, Qi Zhu, Siheng Chen, and Yanfeng Wang. Leapfrog diffusion model for stochastic trajectory prediction. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 5517–5526,

  20. [28]

    Mantra: Memory augmented net- works for multiple trajectory prediction

    Francesco Marchetti, Federico Becattini, Lorenzo Seidenari, and Alberto Del Bimbo. Mantra: Memory augmented net- works for multiple trajectory prediction. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 7143–7152, 2020. 2, 6

    2020

  21. [29]

    Forecast- ing human trajectory from scene history

    Mancheng Meng, Ziyan Wu, Terrence Chen, Xiran Cai, Xiang Zhou, Fan Yang, and Dinggang Shen. Forecast- ing human trajectory from scene history. arXiv preprint arXiv:2210.08732, 2022. 1, 6

    2022 arXiv

  22. [30]

    Dag-net: Double attentive graph neural network for trajectory forecasting

    Alessio Monti, Alessia Bertugli, Simone Calderara, and Rita Cucchiara. Dag-net: Double attentive graph neural network for trajectory forecasting. In 2020 25th international con- ference on pattern recognition (ICPR) , pages 2551–2558. IEEE, 2021. 3

    2020

  23. [31]

    Trajec- tory learning for activity understanding: Unsupervised, mul- tilevel, and long-term adaptive approach

    Brendan Tran Morris and Mohan Manubhai Trivedi. Trajec- tory learning for activity understanding: Unsupervised, mul- tilevel, and long-term adaptive approach. IEEE transactions on pattern analysis and machine intelligence , 33(11):2287– 2301, 2011. 2

    2011

  24. [32]

    You’ll never walk alone: Modeling social behav- ior for multi-target tracking

    Stefano Pellegrini, Andreas Ess, Konrad Schindler, and Luc Van Gool. You’ll never walk alone: Modeling social behav- ior for multi-target tracking. In2009 IEEE 12th international conference on computer vision, pages 261–268. IEEE, 2009. 5

    2009

  25. [33]

    Learning social etiquette: Human tra- jectory prediction in crowded scenes

    Alexandre Robicquet, Amir Sadeghian, Alexandre Alahi, and Silvio Savarese. Learning social etiquette: Human tra- jectory prediction in crowded scenes. In European Confer- ence on Computer Vision (ECCV), page 5, 2016. 5

    2016

  26. [34]

    Trajectron++: Dynamically-feasible tra- jectory forecasting with heterogeneous data

    Tim Salzmann, Boris Ivanovic, Punarjay Chakravarty, and Marco Pavone. Trajectron++: Dynamically-feasible tra- jectory forecasting with heterogeneous data. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XVIII 16, pages...

    2020

  27. [35]

    Sgcn: Sparse graph convolution network for pedestrian trajectory prediction

    Liushuai Shi, Le Wang, Chengjiang Long, Sanping Zhou, Mo Zhou, Zhenxing Niu, and Gang Hua. Sgcn: Sparse graph convolution network for pedestrian trajectory prediction. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 8994–9003, 2021. 3

    2021

  28. [36]

    Pedestrian trajectory prediction based on deep convolutional lstm network

    Xiao Song, Kai Chen, Xu Li, Jinghan Sun, Baocun Hou, Yong Cui, Baochang Zhang, Gang Xiong, and Zilie Wang. Pedestrian trajectory prediction based on deep convolutional lstm network. IEEE Transactions on Intelligent Transporta- tion Systems, 22(6):3285–3302, 2020. 2, 6

    2020

  29. [37]

    Trajectory forecasting based on prior-aware directed graph convolutional neural network

    Yuchao Su, Jie Du, Yuanman Li, Xia Li, Rongqin Liang, Zhongyun Hua, and Jiantao Zhou. Trajectory forecasting based on prior-aware directed graph convolutional neural network. IEEE Transactions on Intelligent Transportation Systems, 23(9):16773–16785, 2022. 3

    2022

  30. [38]

    Stepwise goal-driven networks for trajectory predic- tion

    Chuhua Wang, Yuchen Wang, Mingze Xu, and David J Cran- dall. Stepwise goal-driven networks for trajectory predic- tion. IEEE Robotics and Automation Letters , 7(2):2716– 2723, 2022. 2, 6

    2022

  31. [39]

    View vertically: A hierarchical network for trajectory prediction via fourier spectrums

    Conghao Wong, Beihao Xia, Ziming Hong, Qinmu Peng, Wei Yuan, Qiong Cao, Yibo Yang, and Xinge You. View vertically: A hierarchical network for trajectory prediction via fourier spectrums. In European Conference on Computer Vision, pages 682–700. Springer, 2022. 2, 6

    2022

  32. [40]

    Another vertical view: A hierarchical network for hetero- geneous trajectory prediction via spectrums

    Conghao Wong, Beihao Xia, Qinmu Peng, and Xinge You. Another vertical view: A hierarchical network for hetero- geneous trajectory prediction via spectrums. arXiv preprint arXiv:2304.05106, 2023. 2, 6

    2023 arXiv

  33. [41]

    Socialcircle: Learning the angle-based social in- teraction representation for pedestrian trajectory prediction

    Conghao Wong, Beihao Xia, Ziqian Zou, Yulong Wang, and Xinge You. Socialcircle: Learning the angle-based social in- teraction representation for pedestrian trajectory prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19005–...

    2024

  34. [42]

    Cscnet: Contextual semantic consistency net- work for trajectory prediction in crowded spaces

    Beihao Xia, Conghao Wong, Qinmu Peng, Wei Yuan, and Xinge You. Cscnet: Contextual semantic consistency net- work for trajectory prediction in crowded spaces. Pattern Recognition, 126:108552, 2022. 3

    2022

  35. [43]

    Pedestrian trajectory prediction in heterogeneous traffic us- ing facial keypoints-based convolutional encoder-decoder network

    Song Xiao, Kai Chen, Xiaoxiang Ren, and Haitao Yuan. Pedestrian trajectory prediction in heterogeneous traffic us- ing facial keypoints-based convolutional encoder-decoder network. ACM Transactions on Internet Technology, 22(4): 1–14, 2022. 2

    2022

  36. [44]

    Groupnet: Multiscale hypergraph neural net- works for trajectory prediction with relational reasoning

    Chenxin Xu, Maosen Li, Zhenyang Ni, Ya Zhang, and Si- heng Chen. Groupnet: Multiscale hypergraph neural net- works for trajectory prediction with relational reasoning. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition , pages 6498–6507, 2022. 3, 6

    2022

  37. [45]

    Remember intentions: Retrospective-memory-based trajec- tory prediction

    Chenxin Xu, Weibo Mao, Wenjun Zhang, and Siheng Chen. Remember intentions: Retrospective-memory-based trajec- tory prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 6488– 6497, 2022. 1, 6

    2022

  38. [46]

    Eqmo- tion: Equivariant multi-agent motion prediction with invari- ant interaction reasoning

    Chenxin Xu, Robby T Tan, Yuhong Tan, Siheng Chen, Yu Guang Wang, Xinchao Wang, and Yanfeng Wang. Eqmo- tion: Equivariant multi-agent motion prediction with invari- ant interaction reasoning. arXiv preprint arXiv:2303.10876,

    arXiv

  39. [47]

    Bitrap: Bi-directional pedestrian trajectory prediction with multi-modal goal estimation.IEEE Robotics and Automation Letters, 6(2):1463–1470, 2021

    Yu Yao, Ella Atkins, Matthew Johnson-Roberson, Ram Va- sudevan, and Xiaoxiao Du. Bitrap: Bi-directional pedestrian trajectory prediction with multi-modal goal estimation.IEEE Robotics and Automation Letters, 6(2):1463–1470, 2021. 2, 6

    2021

  40. [48]

    Mul- timodal transformer networks for pedestrian trajectory pre- diction

    Ziyi Yin, Ruijin Liu, Zhiliang Xiong, and Zejian Yuan. Mul- timodal transformer networks for pedestrian trajectory pre- diction. In Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence , pages 1259– 1265, 2021. 2, 3, 6

    2021

  41. [49]

    Spatio-temporal graph transformer networks for pedestrian trajectory prediction

    Cunjun Yu, Xiao Ma, Jiawei Ren, Haiyu Zhao, and Shuai Yi. Spatio-temporal graph transformer networks for pedestrian trajectory prediction. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XII 16, pages 507–523, 2020. 2, 3

    2020

  42. [50]

    Game theoretical interactions of moving agents

    Wenjian Yu and Dirk Helbing. Game theoretical interactions of moving agents. In Simulating Complex Systems by Cellu- lar Automata, pages 219–239. Springer, 2010. 2

    2010

  43. [51]

    Agentformer: Agent-aware transformers for socio-temporal multi-agent forecasting

    Ye Yuan, Xinshuo Weng, Yanglan Ou, and Kris M Kitani. Agentformer: Agent-aware transformers for socio-temporal multi-agent forecasting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 9813– 9823, 2021. 2, 6

    2021

  44. [52]

    Human tra- jectory prediction via neural social physics

    Jiangbei Yue, Dinesh Manocha, and He Wang. Human tra- jectory prediction via neural social physics. In European conference on computer vision, pages 376–394, 2022. 3, 6

    2022

  45. [53]

    Learning fine-grained spatial models for dynamic sports play prediction

    Yisong Yue, Patrick Lucey, Peter Carr, Alina Bialkowski, and Iain Matthews. Learning fine-grained spatial models for dynamic sports play prediction. In 2014 IEEE international conference on data mining, pages 670–679. IEEE, 2014. 5

    2014

  46. [54]

    Sr-lstm: State refinement for lstm to- wards pedestrian trajectory prediction

    Pu Zhang, Wanli Ouyang, Pengfei Zhang, Jianru Xue, and Nanning Zheng. Sr-lstm: State refinement for lstm to- wards pedestrian trajectory prediction. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12085–12094, 2019. 2

    2019

  47. [55]

    Conflict game in evacu- ation process: A study combining cellular automata model

    Xiaoping Zheng and Yuan Cheng. Conflict game in evacu- ation process: A study combining cellular automata model. Physica A: Statistical Mechanics and its Applications , 390 (6):1042–1050, 2011. 2

    2011

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.