Pith. sign in

REVIEW 5 major objections 5 minor 71 references

Learning Group Interactions and Semantic Intentions for Multi-Object Trajectory Prediction

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

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

Pith's one-line read The paper argues that conditioning a diffusion-based trajectory predictor on team-level tactic labels and game-theoretic Banzhaf interaction scores improves multi-agent trajectory and tactic forecasting in team sports.

desk verdict Tactic-conditioned diffusion for sports trajectory prediction is a real empirical idea with a usable new benchmark, but the semantic-intention module overclaims Banzhaf and the headline gains may rest on oracle-like tactic labels. read the letter →

arxiv 2412.15673 v1 pith:FYKK2TI2 submitted 2024-12-20 cs.CV

Mengshi Qi , Yuxin Yang , Huadong Ma This is my paper · ORCID
classification cs.CV
keywords trajectorypredictionmulti-agentforecastingdiffusionmodelclassifier-freeguidancegroupinteractionBanzhaftacticsportsanalytics
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 group-level semantic context improves multi-agent trajectory prediction in team sports. It builds a diffusion-based forecaster whose denoising process is conditioned not only on observed positions but on each team's current tactic label, and it adds a game-theoretic scoring of how strongly agents and tactics cooperate. On the NBA SportVU dataset the method reports the best accuracy among the compared methods, improving on the reproduced LED baseline; it also reports gains on TeamTrack soccer and basketball. If the claim holds, tactic-conditioned diffusion is a workable route to more realistic and interpretable sports trajectory forecasts.

What carries the argument

The carrying object is the interaction-condition vector $G = f_\phi(X, L^{-T_{obs}+1:0})$, where each agent's token is the concatenation of a learned trajectory embedding and a tactic embedding drawn from a vocabulary; the denoising model then follows the classifier-free guidance rule $f_\epsilon(y, G) = f_\epsilon(y, A) + s_g (f_\epsilon(y, G) - f_\epsilon(y, A))$. The second mechanism is Banzhaf Interaction, a cooperative-game score for how much a coalition's joint contribution exceeds its members' separate contributions, computed between agents and top-$k$ candidate tactics; its predicted values are fused with globally and locally attended agent tokens, and a graph-attention head outputs tactic probabilities.

What would settle it

Permute or randomly drop the observed tactic labels at inference on NBA SportVU and re-measure minADE/minFDE: if scores stay near 0.19/0.29 rather than reverting toward the LED baseline's 0.21/0.31, the tactic condition is not the cause of the gain. Re-running TeamTrack evaluation with independent tactic ground truth instead of the K-means pseudo-labels would test whether those labels carry the improvement.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that group-level semantic context is a usable condition for trajectory generation. The authors construct the condition vector $G = f_\phi(X, L^{-T_{obs}+1:0})$ by concatenating each agent's trajectory embedding with an embedding of its team's current tactic, feed it into the denoising network through classifier-free guidance, and supervise an auxiliary tactic predictor with Banzhaf Interaction scores computed between agents and candidate tactics. They report minADE/minFDE of 0.19/0.29 at 1s and 0.84/1.17 at 4s on NBA SportVU, down from the reproduced LED baseline's 0.21/0.31 and 0.89/1.24, and Top-1 tactic accuracy of 60.23%; on TeamTrack soccer and basketball the same recipe also improves on the compared baselines.

Load-bearing premise

The load-bearing premise is that tactic labels for the observed frames are available and reliable at inference, since on the TeamTrack datasets the labels come from K-means clustering without external ground truth and any noise or bias in those labels could erase the reported trajectory gains.

Editorial extensions

If this is right

  • Tactic-conditioned diffusion gives a template for group-behavior forecasting wherever a team-level semantic label is available.
  • One forward pass produces both future trajectories and future tactic probabilities, so a downstream system could anticipate an offensive or defensive switch before it is visible in player positions.
  • Because the method uses classifier-free guidance, tactic information can be randomly dropped during training, letting users tune guidance strength at inference to balance diversity and fidelity.
  • The manually annotated NBA tactic benchmark provides a supervised target for future tactic-conditioned trajectory models and for studying label noise.

Reading between the lines

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

  • Beyond the paper, a learned tactic recognizer could replace the observed labels at inference, since the current formulation consumes manual or pseudo labels as input rather than detecting them from motion alone.
  • Beyond the paper, the computed Banzhaf Interaction values could serve as a per-player importance map accompanying each forecast, extending the paper's visualization that the ball handler and nearest defenders receive the highest scores.
  • Beyond the paper, corrupting tactic labels with controlled noise and measuring minADE/minFDE would map how accurate an automatic labeler must be for the reported gains to survive.
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 a diffusion-based trajectory prediction framework for team sports that conditions trajectory generation on observed group-level tactic labels, enhances agent embeddings with global/local attention, and introduces an auxiliary tactic-prediction task supervised by a learned 'Banzhaf Interaction' module. The authors extend the NBA SportVU dataset with human annotations of 16 team tactics and generate K-means pseudo-labels for the TeamTrack datasets. Experiments report state-of-the-art minADE/minFDE on NBA SportVU and TeamTrack, as well as improved Top-k tactic prediction accuracy over Transformer, pooling, and LLM baselines. The central claim is that group-level semantic context, modeled via game-theoretic interaction, improves both trajectory and tactic forecasting.

Significance. If validated, the framework would be a meaningful step toward incorporating group-level semantic knowledge into generative trajectory prediction, and the human-annotated NBA SportVU tactic dataset is a useful community resource. The paper ships code and data, reproduces LED and other baselines, and evaluates against LLM baselines, which is commendable. However, the significance is contingent on resolving several load-bearing concerns: the annotation protocol may permit future information leakage into the conditioning signal; the 'Banzhaf Interaction' computation in Eq. (24) does not implement the game-theoretic definition in Eq. (5) and its supervision is generated from the model's own predictions; the TeamTrack pseudo-labels lack external validation; and the reported numeric gains are presented without error bars or a condition-ablation that isolates the contribution of tactic labels.

major comments (5)
  1. [Sec. 5.2 / Eq. (8)] The paper never states whether human annotators assigned past tactic labels L^{-T_obs+1:0} using only the observed 10 frames or the full 30-frame clip. If annotators watched the future 20 frames, the conditioning variable in Eq. (8) encodes future trajectory information, directly inflating the Table 1 gains over label-free baselines (LED, MID, SingularTrajectory). The authors must clarify the annotation protocol and, if necessary, re-annotate or demonstrate that labels are assigned from observed history alone; otherwise the central trajectory claim is not attributable to the method.
  2. [Sec. 4.5.3, Eq. (24) vs Eq. (5)] Eq. (24) defines I^j_B as the average of two masked softmax logits (Sa2t and St2a), which is not the Banzhaf interaction defined in Eq. (5): there is no summation over coalitions C with p(C)=1/2^{n-2} and no characteristic-function payoff differences. Additionally, the supervision I^j_B is computed from the model's own predicted Top-k tactic embeddings c_hat_j (Sec. 4.5.2), making the 'ground truth' self-generated rather than an externally anchored game-theoretic value. This undermines the paper's claim to model semantic intentions via Banzhaf interaction; either implement the actual Banzhaf interaction or rename the module and reframe the claims.
  3. [Tables 1, 2, 3] No error bars, standard deviations, or statistical significance tests are reported for any of the trajectory or tactic prediction results. The claimed improvements over LED (e.g., 0.19 vs 0.21 minADE at 1s in Table 1) may be within run-to-run variance. The authors should report means and standard deviations over multiple random seeds and, if possible, perform significance testing to support the state-of-the-art claim.
  4. [Sec. 5.2 / Table 2] The TeamTrack pseudo-labels are produced by K-means clustering of trajectories into 16 clusters, with no external validation that these clusters correspond to meaningful team tactics. The t-SNE visualization in Fig. 3 and the referenced generalizability experiments do not establish such correspondence. Without a comparison to human-annotated tactics or other external grounding, the Table 2 results may reflect the model exploiting cluster-ID artifacts rather than semantic intentions. Provide quantitative validation of the pseudo-labels or remove/qualify the TeamTrack claims.
  5. [Sec. 4.2 and Sec. 6.7] The main trajectory improvement is attributed to conditioning on group-level tactic labels L, but no ablation isolates this factor. Table 4 varies the generation backbone (MLP, query, diffusion) while keeping the interaction encoder fixed, and the ablation in Sec. 6.7 does not compare against a diffusion model trained without the tactic condition or with random tactic labels. Without such an ablation, the reported gains cannot be attributed to the semantic intention mechanism rather than to architectural choices.
minor comments (5)
  1. [Sec. 6.3] The implementation details state 'we set the game factor to 0.001 and the γ in Eq. (21) to 4.0. The tactic prediction rate is set at 1000.' It is unclear which symbols correspond to α, β, and η in Eq. (25), and what 'tactic prediction rate' means; please clarify.
  2. [Table 5] The BI module improves Top-1 accuracy by only about 1.8 percentage points (from 58.44 to 60.23); given the absence of error bars, it is unclear whether this gain is statistically significant.
  3. [Eq. (9)] The proportionality '∝' in the classifier-free guidance update is loose; the exact update is fϵ(y, G) = fϵ(y, A) + sg·(fϵ(y, G) − fϵ(y, A)). Please state the update directly.
  4. [Sec. 4.5.2, Eq. (18)] The notation 'arg max_k (p_j)' is ambiguous; it should be defined as selecting the indices of the k largest probabilities, not as a standard argmax over a single value.
  5. [Fig. 3] The t-SNE plot alone does not validate that K-means clusters align with semantic tactic categories; please add a quantitative measure (e.g., cluster–label agreement on a human-annotated subset) or move the figure to supplementary with explicit caveats.

Circularity Check

2 steps flagged · score 6.0 of 10

The Banzhaf 'ground-truth' interaction is generated by the model's own predicted tactic embeddings, and the tactic-condition input may be annotated from full clips containing the future, so the semantic-intention mechanism and the Table 1 gain are not yet established as causal.

  1. fitted input called prediction [Sec. 4.5.2-4.5.3, Eq. (15) and Eq. (24)]
    "This part takes the similarity matrix of agent tokens for each group and the predicted Top-k tactic embeddings ĉj retrieved from Sec. 4.5.2 as inputs, which can be viewed as the interaction logits S... The output can be seen as the ground-truth of Banzhaf Interaction, denoted as IB. ... L j bi = ‖ I j pred − I j B ‖ 2 , Lbi = P M j=1 L j bi."

    The 'ground-truth' IB used to supervise the Banzhaf Interaction Learner is computed by Eq. (24) from S, and S is constructed from ĉj, the model's own predicted Top-k tactic embeddings from Eq. (18). The learner is therefore trained to match a function of the very predictions it is supposed to produce; no external measurement of Banzhaf interaction enters the loop. This makes the semantic-intention module a self-consistency loss rather than an estimator of the Banzhaf interaction defined in Eq. (5).

  2. self definitional [Sec. 5.2 annotation protocol with Sec. 4.2 Eq. (8) and Eq. (11)]
    "We construct the extended NBA SportVU dataset, annotated with human-labeled tactics for the previous 10 frames and the following 20 frames at 5 Hz. Annotators select one tactic for each team and time period from 16 representative team tactics. ... gi = [ ai; ci,m(i) ] ... Thus, we define the function of Interaction Encoder as G = fϕ(X, L−Tobs+1:0)."

    The condition G = fϕ(X, L−Tobs+1:0) uses L−Tobs+1:0, the past tactic label, as an input to predict the future trajectories X1:Tpred. The annotation protocol labels 'the previous 10 frames and the following 20 frames' of the same clip, and the paper does not state that the past-tactic label was assigned using only the 10 observed frames. If annotators used the full 30-frame clip, the input label encodes future trajectory information, so the reported minADE/minFDE advantage over baselines that receive no tactic condition is a label-leakage artifact rather than evidence for the semantic-intention mechanism.

full rationale

The core trajectory and tactic losses are not wholly self-referential: Ldist and Ltactic are anchored to ground-truth future positions and human-annotated future tactic labels, respectively, and the diffusion denoising follows the standard conditional diffusion objective. However, two load-bearing parts of the derivation are circular or potentially leaky. First, the Banzhaf Interaction Learner's supervision IB is generated from the model's own predicted Top-k tactic embeddings, so the claimed game-theoretic 'ground truth' is a function of the model's own outputs; the ablation in Table 5 therefore does not demonstrate that external cooperation semantics improve prediction. Second, the NBA tactic labels used as conditioning may have been annotated with access to the full clip including the future frames being predicted, and the paper does not establish that the observed-window labels are causally prior to the prediction target. On TeamTrack, the K-means pseudo-labels add a further uncertainty: they are cluster IDs with no external tactic ground truth, and the paper does not specify whether clustering used only observed motion or full sequences. These issues mean the headline trajectory improvements are not yet attributable to the proposed semantic-intention mechanism, although the diffusion and tactic-classification components retain independent empirical content. For that reason the paper is partially circular rather than fully so.

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

The central claim rests on the availability and validity of tactic labels, on the standard diffusion framework, and on the unsupported identification of Eq (24) with the game-theoretic Banzhaf Interaction. The free parameters are mostly unstated or hand-set, and the semantic-intention "ground truth" is generated by the model itself.

free parameters (5)
  • Classifier-free guidance scale s_g = 0.1
    Set by hand in Sec 6.3; directly controls how strongly tactic conditioning influences generated trajectories in Eq (9).
  • Loss weights eta, alpha, beta = not fully reported
    Used in the total objective Eq (25); only "game factor 0.001" and "tactic prediction rate 1000" are mentioned, with no explicit values for eta, alpha, and beta.
  • Focal loss gamma = 4.0
    Set in Sec 6.3 for the tactic focal loss in Eq (20).
  • Top-k potential tactics k = not reported
    The Banzhaf Interaction Learner outputs an N by k matrix and the tactic head extracts Top-k labels (Sec 4.5), but no value or search procedure is reported.
  • Number of K-means clusters = 16
    Sec 5.2 uses 16 clusters patterned on the NBA tactic taxonomy to generate pseudo-labels for TeamTrack, with no validation against real tactic labels.
assumptions (6)
  • standard math DDPM forward and reverse Markov chains with a fixed noise schedule correctly model future trajectories.
    Sec 3.1 Eqs (1) to (4) reuse the standard DDPM formulation from [13,46].
  • domain assumption Classifier-free guidance can be applied to trajectory generation by randomly dropping tactic embeddings during training.
    Eq (9) assumes the guidance approximation from [17] transfers from image generation to multi-agent trajectory prediction.
  • domain assumption The 16-tactic vocabulary covers the meaningful team strategies in NBA basketball.
    Sec 5.1 selects 16 tactics by surveying videos and coaching manuals, but no coverage analysis or inter-annotator agreement is reported.
  • ad hoc to paper K-means clustering of trajectories produces valid tactic pseudo-labels for TeamTrack.
    Sec 5.2 generates pseudo-labels to resemble the manual annotation scheme and uses them both as training conditions and as validation, without external tactic ground truth.
  • ad hoc to paper The masked-softmax average in Eq (24) equals the Banzhaf Interaction defined in Eq (5).
    Sec 4.5.3 labels I_B as Banzhaf Interaction, but Eqs (22) to (24) compute a softmax average of agent-tactic logits, not the coalition sum in Eq (5).
  • domain assumption Human annotations on the NBA SportVU extension are accurate and consistent.
    Sec 5.2 describes training annotators and manual verification, but no agreement statistics or reliability protocol are reported.
invented entities (1)
  • Semantic intention score I_B per agent-tactic pair
    purpose: Used as learned pseudo-ground-truth to supervise the Banzhaf Interaction Learner and as a fusion weight for tactic prediction.
    There is no external measure of "semantic intention"; I_B is computed from the model's own predicted tactic embeddings c_hat_j (Sec 4.5.2 and 4.5.3), so it is internal to the model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning Group Interactions and Semantic Intentions for Multi-Object Trajectory Prediction." pith.science (2026). https://pith.science/paper/FYKK2TI2

@misc{pith2026241215673,
  author       = {Pith},
  title        = {Pith review of: Learning Group Interactions and Semantic Intentions for Multi-Object Trajectory Prediction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FYKK2TI2}},
  note         = {Machine review of arXiv:2412.15673}
}
read the original abstract

Effective modeling of group interactions and dynamic semantic intentions is crucial for forecasting behaviors like trajectories or movements. In complex scenarios like sports, agents' trajectories are influenced by group interactions and intentions, including team strategies and opponent actions. To this end, we propose a novel diffusion-based trajectory prediction framework that integrates group-level interactions into a conditional diffusion model, enabling the generation of diverse trajectories aligned with specific group activity. To capture dynamic semantic intentions, we frame group interaction prediction as a cooperative game, using Banzhaf interaction to model cooperation trends. We then fuse semantic intentions with enhanced agent embeddings, which are refined through both global and local aggregation. Furthermore, we expand the NBA SportVU dataset by adding human annotations of team-level tactics for trajectory and tactic prediction tasks. Extensive experiments on three widely-adopted datasets demonstrate that our model outperforms state-of-the-art methods. Our source code and data are available at https://github.com/aurora-xin/Group2Int-trajectory.

Figures

Figures reproduced from arXiv: 2412.15673 by the authors.

Figure 1
Figure 1. Illustration of multi-agent trajectory prediction in a 3 vs. 3 basket [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview of our proposed method. It consists of two main parts: (1) denoising module for diffusion-based trajectory prediction and (2) [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. 3D t-SNE visualized results of utilizing clustering to generate [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Illustration of the team-level tactic annotation distribution of NBA [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Illustration of visualization results of trajectory prediction on the NBA SportVU dataset, where the ground truth show in the last column, LED [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Qualitative results of two-fold prediction tasks on NBA SportVU [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: The visualization results of the predicted Banzhaf Interaction on the NBA SportVU dataset. This figure shows the ground truth (in the last [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 61 canonical work pages

  1. [1]

    Where will they go? predict- ing fine-grained adversarial multi-agent motion using conditional variational autoencoders,

    P . Felsen, P . Lucey, and S. Ganguly, “Where will they go? predict- ing fine-grained adversarial multi-agent motion using conditional variational autoencoders,” in Proceedings of the European conference on computer vision, 2018, pp. 732–747

    work page 2018

  2. [2]

    Imitative non-autoregressive modeling for trajectory forecasting and imputation,

    M. Qi, J. Qin, Y. Wu, and Y. Yang, “Imitative non-autoregressive modeling for trajectory forecasting and imputation,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2020, pp. 12 736–12 745

    work page 2020

  3. [3]

    stagnet: An attentive semantic rnn for group activity recognition,

    M. Qi, J. Qin, A. Li, Y. Wang, J. Luo, and L. Van Gool, “stagnet: An attentive semantic rnn for group activity recognition,” in Proceedings of the European conference on computer vision , 2018, pp. 101–117

    work page 2018

  4. [4]

    Groupnet: Multi- scale hypergraph neural networks for trajectory prediction with relational reasoning,

    C. Xu, M. Li, Z. Ni, Y. Zhang, and S. Chen, “Groupnet: Multi- scale hypergraph neural networks for trajectory prediction with relational reasoning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6498–6507

    work page 2022

  5. [5]

    Query-centric trajectory prediction,

    Z. Zhou, J. Wang, Y.-H. Li, and Y.-K. Huang, “Query-centric trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 17 863–17 873

    work page 2023

  6. [6]

    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/CVF Conference on Computer Vision and Pattern Recognition, 2016, pp. 961–971

    work page 2016

  7. [7]

    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 Conference on Computer Vision and Pattern Recognition, 2019, pp. 6272–6281

    work page 2019

  8. [8]

    Bifold and semantic reasoning for pedestrian behavior prediction,

    A. Rasouli, M. Rohani, and J. Luo, “Bifold and semantic reasoning for pedestrian behavior prediction,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 15 600–15 610. 13

    work page 2021

Show all 71 references
  1. [9]

    Towards efficient human-robot collaboration with robust plan recognition and tra- jectory prediction,

    Y. Cheng, L. Sun, C. Liu, and M. Tomizuka, “Towards efficient human-robot collaboration with robust plan recognition and tra- jectory prediction,” IEEE Robotics and Automation Letters , vol. 5, no. 2, pp. 2602–2609, 2020

    2020

  2. [10]

    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/CVF Conference on Computer Vision and Pattern Recognition, 2018, pp. 2255–2264

    2018

  3. [11]

    Trajec- tron++: Dynamically-feasible trajectory forecasting with heteroge- neous data,

    T. Salzmann, B. Ivanovic, P . Chakravarty, and M. Pavone, “Trajec- tron++: Dynamically-feasible trajectory forecasting with heteroge- neous data,” in Proceedings of the European conference on computer vision. Springer, 2020, pp. 683–700

    2020

  4. [12]

    It is not the journey but the destination: Endpoint conditioned trajectory prediction,

    K. Mangalam, H. Girase, S. Agarwal, K.-H. Lee, E. Adeli, J. Malik, and A. Gaidon, “It is not the journey but the destination: Endpoint conditioned trajectory prediction,” in Proceedings of the European conference on computer vision. Springer, 2020, pp. 759–776

    2020

  5. [13]

    Stochastic trajectory prediction via motion indeterminacy diffusion,

    T. Gu, G. Chen, J. Li, C. Lin, Y. Rao, J. Zhou, and J. Lu, “Stochastic trajectory prediction via motion indeterminacy diffusion,” in Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 17 113–17 122

    2022

  6. [14]

    Leapfrog diffusion model for stochastic trajectory prediction,

    W. Mao, C. Xu, Q. Zhu, S. Chen, and Y. Wang, “Leapfrog diffusion model for stochastic trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 5517–5526

    2023

  7. [15]

    Singulartrajectory: Universal trajectory predictor using diffusion model,

    I. Bae, Y.-J. Park, and H.-G. Jeon, “Singulartrajectory: Universal trajectory predictor using diffusion model,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2024, pp. 17 890–17 901

    2024

  8. [16]

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

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

    2020

  9. [17]

    Classifier-free diffusion guidance,

    J. Ho and T. Salimans, “Classifier-free diffusion guidance,” CoRR, vol. abs/2207.12598, 2022

    2022 arXiv

  10. [18]

    Video-text as game players: Hierarchical banzhaf interac- tion for cross-modal representation learning,

    P . Jin, J. Huang, P . Xiong, S. Tian, C. Liu, X. Ji, L. Yuan, and J. Chen, “Video-text as game players: Hierarchical banzhaf interac- tion for cross-modal representation learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp...

    2023

  11. [19]

    Ask me anything: Dynamic memory networks for natural language processing,

    A. Kumar, O. Irsoy, P . Ondruska, M. Iyyer, J. Bradbury, I. Gulrajani, V . Zhong, R. Paulus, and R. Socher, “Ask me anything: Dynamic memory networks for natural language processing,” in Interna- tional Conference on Machine Learning . Proceedings of Machine Learning Research,...

    2016

  12. [20]

    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

    1995

  13. [21]

    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

    2018

  14. [22]

    Stc-gan: Spatio-temporally coupled generative adversarial networks for predictive scene pars- ing,

    M. Qi, Y. Wang, A. Li, and J. Luo, “Stc-gan: Spatio-temporally coupled generative adversarial networks for predictive scene pars- ing,” IEEE Transactions on Image Processing, vol. 29, pp. 5420–5430, 2020

    2020

  15. [23]

    Weakly-supervised temporal action localization by inferring salient snippet-feature,

    W. Yun, M. Qi, C. Wang, and H. Ma, “Weakly-supervised temporal action localization by inferring salient snippet-feature,” in Proceed- ings of the AAAI Conference on Artificial Intelligence , vol. 38, no. 7, 2024, pp. 6908–6916

    2024

  16. [24]

    Attentive relational networks for mapping images to scene graphs,

    M. Qi, W. Li, Z. Yang, Y. Wang, and J. Luo, “Attentive relational networks for mapping images to scene graphs,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019, pp. 3957–3966

    2019

  17. [25]

    Evolvegraph: Multi- agent trajectory prediction with dynamic relational reasoning,

    J. Li, F. Yang, M. Tomizuka, and C. Choi, “Evolvegraph: Multi- agent trajectory prediction with dynamic relational reasoning,” Advances in neural information processing systems, vol. 33, pp. 19 783– 19 794, 2020

    2020

  18. [26]

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

    A. Mohamed, D. Zhu, W. Vu, M. Elhoseiny, and C. Claudel, “Social-implicit: Rethinking trajectory prediction evaluation and the effectiveness of implicit maximum likelihood estimation,” in Proceedings of the European conference on computer vision . Springer, 2022, pp. 463–479

    2022

  19. [27]

    Sgcn: Sparse graph convolution network for pedestrian trajectory prediction,

    L. Shi, L. Wang, C. Long, S. Zhou, M. Zhou, Z. Niu, and G. Hua, “Sgcn: Sparse graph convolution network for pedestrian trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 8994–9003

    2021

  20. [28]

    Adaptive trajectory prediction via transferable gnn,

    Y. Xu, L. Wang, Y. Wang, and Y. Fu, “Adaptive trajectory prediction via transferable gnn,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 6520–6531

    2022

  21. [29]

    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. 1349–1358

    2019

  22. [30]

    Mg-gan: A multi- generator model preventing out-of-distribution samples in pedes- trian trajectory prediction,

    P . Dendorfer, S. Elflein, and L. Leal-Taix ´e, “Mg-gan: A multi- generator model preventing out-of-distribution samples in pedes- trian trajectory prediction,” in Proceedings of the IEEE/CVF Interna- tional Conference on Computer Vision, 2021, pp. 13 158–13 167

    2021

  23. [31]

    Stimulus verification is a universal and effective sampler in multi-modal human trajectory prediction,

    J. Sun, Y. Li, L. Chai, and C. Lu, “Stimulus verification is a universal and effective sampler in multi-modal human trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 22 014–22 023

    2023

  24. [32]

    Personalized trajectory prediction via distribution discrimination,

    G. Chen, J. Li, N. Zhou, L. Ren, and J. Lu, “Personalized trajectory prediction via distribution discrimination,” in Proceedings of the IEEE/CVF International Conference on Computer Vision , 2021, pp. 15 580–15 589

    2021

  25. [33]

    The trajectron: Probabilistic multi- agent trajectory modeling with dynamic spatiotemporal graphs,

    B. Ivanovic and M. Pavone, “The trajectron: Probabilistic multi- agent trajectory modeling with dynamic spatiotemporal graphs,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019, pp. 2375–2384

    2019

  26. [34]

    Muse-vae: Multi-scale vae for environment-aware long term tra- jectory prediction,

    M. Lee, S. S. Sohn, S. Moon, S. Yoon, M. Kapadia, and V . Pavlovic, “Muse-vae: Multi-scale vae for environment-aware long term tra- jectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 2221–2230

    2022

  27. [35]

    Desire: Distant future prediction in dynamic scenes with interacting agents,

    N. Lee, W. Choi, P . Vernaza, C. B. Choy, P . H. Torr, and M. Chan- draker, “Desire: Distant future prediction in dynamic scenes with interacting agents,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2017, pp. 336–345

    2017

  28. [36]

    Trace and pace: Controllable pedestrian animation via guided trajectory diffusion,

    D. Rempe, Z. Luo, X. Bin Peng, Y. Yuan, K. Kitani, K. Kreis, S. Fidler, and O. Litany, “Trace and pace: Controllable pedestrian animation via guided trajectory diffusion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 13 756–13 766

    2023

  29. [37]

    Motiondiffuser: Controllable multi-agent motion prediction using diffusion,

    C. Jiang, A. Cornman, C. Park, B. Sapp, Y. Zhou, D. Anguelov et al., “Motiondiffuser: Controllable multi-agent motion prediction using diffusion,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 9644–9653

    2023

  30. [38]

    Following in the footsteps: Predicting human trajectories using motion pattern memory,

    Y. Yang, P . Zhu, M. Qi, and H. Ma, “Following in the footsteps: Predicting human trajectories using motion pattern memory,” in Proceedings of the 6th ACM International Conference on Multimedia in Asia, 2024

    2024

  31. [39]

    Ss-lstm: A hierarchical lstm model for pedestrian trajectory prediction,

    H. Xue, D. Q. Huynh, and M. Reynolds, “Ss-lstm: A hierarchical lstm model for pedestrian trajectory prediction,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision . IEEE, 2018, pp. 1186–1194

    2018

  32. [40]

    Mantra: Memory augmented networks for multiple trajectory prediction,

    F. Marchetti, F. Becattini, L. Seidenari, and A. D. Bimbo, “Mantra: Memory augmented networks for multiple trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 7143–7152

    2020

  33. [41]

    Ball trajectory inference from multi-agent sports contexts using set transformer and hierarchical bi-lstm,

    H. Kim, H.-J. Choi, C. J. Kim, J. Yoon, and S.-K. Ko, “Ball trajectory inference from multi-agent sports contexts using set transformer and hierarchical bi-lstm,” in Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023, pp. 4296– 4307

    2023

  34. [42]

    Multi-modal trajectory prediction of nba players,

    S. Hauri, N. Djuric, V . Radosavljevic, and S. Vucetic, “Multi-modal trajectory prediction of nba players,” inProceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021, pp. 1640– 1649

    2021

  35. [43]

    Sports video captioning via attentive motion representation and group relationship model- ing,

    M. Qi, Y. Wang, A. Li, and J. Luo, “Sports video captioning via attentive motion representation and group relationship model- ing,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 30, no. 8, pp. 2617–2633, 2019

    2019

  36. [44]

    Semi-supervised teacher- reference-student architecture for action quality assessment,

    W. Yun, M. Qi, F. Peng, and H. Ma, “Semi-supervised teacher- reference-student architecture for action quality assessment,” in Proceedings of the European conference on computer vision . Springer, 2025, pp. 161–178

    2025

  37. [45]

    Semantics-aware spatial-temporal binaries for cross-modal video retrieval,

    M. Qi, J. Qin, Y. Yang, Y. Wang, and J. Luo, “Semantics-aware spatial-temporal binaries for cross-modal video retrieval,” IEEE Transactions on Image Processing, vol. 30, pp. 2989–3004, 2021

    2021

  38. [46]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P . Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840–6851, 2020

    2020

  39. [47]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” arXiv preprint arXiv:2010.02502, 2020. 14

    2010 arXiv

  40. [48]

    Structured denoising diffusion models in discrete state-spaces,

    J. Austin, D. D. Johnson, J. Ho, D. Tarlow, and R. Van Den Berg, “Structured denoising diffusion models in discrete state-spaces,” Advances in Neural Information Processing Systems , vol. 34, pp. 17 981–17 993, 2021

    2021

  41. [49]

    Diffusion models beat gans on image synthesis,

    P . Dhariwal and A. Nichol, “Diffusion models beat gans on image synthesis,” Advances in neural information processing systems, vol. 34, pp. 8780–8794, 2021

    2021

  42. [50]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P . Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 10 684–10 695

    2022

  43. [51]

    Is conditional generative modeling all you need for decision-making?

    A. Ajay, Y. Du, A. Gupta, J. Tenenbaum, T. Jaakkola, and P . Agrawal, “Is conditional generative modeling all you need for decision-making?” arXiv preprint arXiv:2211.15657, 2022

    2022 arXiv

  44. [52]

    Rabin et al., Incorporating fairness into game theory and economics

    M. Rabin et al., Incorporating fairness into game theory and economics. University of California at Berkeley, Department of Economics, 1992

    1992

  45. [53]

    D. M. Kreps, Game theory and economic modelling. Oxford Univer- sity Press, 1990

    1990

  46. [54]

    S. J. Brams, Game theory and politics. Courier Corporation, 2011

    2011

  47. [55]

    Osborne and A

    M. Osborne and A. Rubinstein, A Course in Game Theory , ser. A Course in Game Theory. MIT Press, 1994. [Online]. Available: https://books.google.com.hk/books?id=5ntdaYX4LPkC

    1994

  48. [56]

    Forecast- ing interactive dynamics of pedestrians with fictitious play,

    W.-C. Ma, D.-A. Huang, N. Lee, and K. M. Kitani, “Forecast- ing interactive dynamics of pedestrians with fictitious play,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017, pp. 774–782

    2017

  49. [57]

    Fine-grained semantically aligned vision-language pre- training,

    J. Li, X. He, L. Wei, L. Qian, L. Zhu, L. Xie, Y. Zhuang, Q. Tian, and S. Tang, “Fine-grained semantically aligned vision-language pre- training,” Advances in neural information processing systems , vol. 35, pp. 7290–7303, 2022

    2022

  50. [58]

    A value for n-person games,

    L. S. Shapley, “A value for n-person games,” in Contributions to the Theory of Games II, H. W. Kuhn and A. W. Tucker, Eds. Princeton: Princeton University Press, 1953, pp. 307–317

    1953

  51. [59]

    The shapley value,

    E. Winter, “The shapley value,” ser. Handbook of Game Theory with Economic Applications. Elsevier, 2002, vol. 3, pp. 2025–2054

    2002

  52. [60]

    Weighted voting doesn’t work: A mathematical anal- ysis,

    J. Banzhaf, “Weighted voting doesn’t work: A mathematical anal- ysis,” Rutgers Law Review, vol. 19, no. 2, pp. 317–343, 1965

    1965

  53. [61]

    An axiomatization of the banzhaf value,

    E. Lehrer, “An axiomatization of the banzhaf value,” International Journal of Game Theory, vol. 17, pp. 89–99, 1988

    1988

  54. [62]

    Explaining deep convolutional neural networks via latent visual-semantic filter attention,

    Y. Yang, S. Kim, and J. Joo, “Explaining deep convolutional neural networks via latent visual-semantic filter attention,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion, 2022, pp. 8333–8343

    2022

  55. [63]

    Gameformer: Game-theoretic mod- eling and learning of transformer-based interactive prediction and planning for autonomous driving,

    Z. Huang, H. Liu, and C. Lv, “Gameformer: Game-theoretic mod- eling and learning of transformer-based interactive prediction and planning for autonomous driving,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 3903–3913

    2023

  56. [64]

    Hier- archical text-conditional image generation with clip latents,

    A. Ramesh, P . Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hier- archical text-conditional image generation with clip latents,” arXiv preprint arXiv:2204.06125, vol. 1, no. 2, p. 3, 2022

    2022 arXiv

  57. [65]

    An axiomatic approach to the concept of interaction among players in cooperative games,

    M. Grabisch and M. Roubens, “An axiomatic approach to the concept of interaction among players in cooperative games,” In- ternational Journal of game theory, vol. 28, pp. 547–565, 1999

    1999

  58. [66]

    Focal loss for dense object detection,

    T.-Y. Ross and G. Doll ´ar, “Focal loss for dense object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2017, pp. 2980–2988

    2017

  59. [67]

    Knowledge-inspired 3d scene graph prediction in point cloud,

    S. Zhang, A. Hao, H. Qin et al. , “Knowledge-inspired 3d scene graph prediction in point cloud,” Advances in Neural Information Processing Systems, vol. 34, pp. 18 620–18 632, 2021

    2021

  60. [68]

    Teamtrack: A dataset for multi-sport multi-object tracking in full- pitch videos,

    A. Scott, I. Uchida, N. Ding, R. Umemoto, R. P . Bunker, R. Kobayashi, T. Koyama, M. Onishi, Y. Kameda, and K. Fujii, “Teamtrack: A dataset for multi-sport multi-object tracking in full- pitch videos,” CoRR, vol. abs/2404.13868, 2024

    2024 arXiv

  61. [69]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. K ¨opf, E. Z. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chil- amkurthy, B. Steiner, L. Fang, J. Bai, and S. Chintala, “Pytorch: An imperative style, ...

    2019

  62. [70]

    Socialcircle: Learn- ing the angle-based social interaction representation for pedestrian trajectory prediction,

    C. Wong, B. Xia, Z. Zou, Y. Wang, and X. You, “Socialcircle: Learn- ing the angle-based social interaction representation for pedestrian trajectory prediction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 19 005–19 015

    2024

  63. [71]

    Lora: Low-rank adaptation of large language models,

    E. J. Hu, Y. Shen, P . Wallis, Z. Allen-Zhu, Y. Li, S. Wang, L. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” arXiv preprint arXiv:2106.09685, 2021

    2021 arXiv

Pith tools

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