Orchestrating the Symphony of Prompt Distribution Learning for Human-Object Interaction Detection

Ge Li; Liming Zhao; Mingda Jia; Yun Zheng

REVIEW 2 major objections 6 minor 1 cited by

Orchestrating the Symphony of Prompt Distribution Learning for Human-Object Interaction Detection

T0 review · 2 major / 6 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 Modeling each interaction as a distribution beats fixed queries

desk verdict Solid HOI gains from prompt distribution learning, but the dynamic orthogonal loss as written pulls categories together rather than apart, and the paper ships no code or error bars. read the letter →

arxiv 2412.08506 v1 pith:3BEBGSJW submitted 2024-12-11 cs.CV

Mingda Jia , Liming Zhao , Ge Li , Yun Zheng This is my paper · ORCID

classification cs.CV

keywords human-objectinteractiondetectionpromptdistributionlearningsoftpromptscategoryqueriesGaussianreparameterizationsamplingzero-shotHOItransformer-based

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 argues that a major reason transformer-based HOI detectors miss rare or ambiguous human-object interactions is that each interaction category is represented by a single query or prompt, which cannot capture the many visual forms one action can take. To fix this, the authors introduce InterProDa, which learns several soft prompts per category, estimates a Gaussian distribution over them, and samples from that distribution to form category-aware detector queries. With this change, a single category can express a near-infinite range of visual patterns, and distinct categories can be pushed apart in a shared distribution space. On the standard benchmarks HICO-DET and V-COCO, the approach reaches 42.67 full mAP, 45.21 rare mAP, and 67.6 scenario-1 AP, outperforming prior methods and boosting existing detectors when plugged in with only a few extra parameters.

What carries the argument

The central object is the category distribution query: each HOI category is represented not by a fixed prompt embedding but by a Gaussian distribution estimated from $K$ learnable soft prompts, with mean $\mu$ and variance $\sigma$ in the CLIP text-embedding space. Before feeding queries to the detector, the distribution is sampled via a VAE-style reparameterization with a learnable noise scale $\gamma$, producing a finite set of samples $S(\hat{P}_c)=\hat{\mu}_c+\gamma n \hat{\sigma}_c$ that are linearly projected and added to the decoder queries. This gives each query a continuous range of visual patterns, while a dynamic orthogonal loss on the mean embeddings keeps ambiguous categories separated in the shared distribution space.

What would settle it

Sample many random prompts per HOI category, embed them with the CLIP text encoder, and apply a multivariate normality test (e.g., Henze-Zirkler) to each per-category cloud; if the clouds are clearly multi-modal or heavily skewed, the Gaussian assumption underlying InterProDa's mean/variance estimates fails, undermining the claimed fidelity of the learned distributions.

Watch

Extended reading notes

Core claim

InterProDa claims that HOI category knowledge is better represented as a distribution than as a point. For each HOI category, the method learns a collection of K soft prompt embeddings, assumes these embeddings follow a Gaussian, and estimates their mean and variance. These per-category Gaussians are stacked into distribution spaces for subjects, objects, and interactions, trained jointly with a dynamic orthogonal loss that keeps semantically distinct categories apart. During inference, the distribution space is sampled using a reparameterization trick with a learnable noise scale, and the samples are fused into the decoder queries of a transformer-based HOI detector. The result, the authors report, is a detector that captures diverse intra-category visual patterns and inter-category dependencies, achieving state-of-the-art performance on HICO-DET and V-COCO and improving existing one-stage and two-stage detectors with small parameter overhead.

Load-bearing premise

The method assumes that all prompt embeddings describing one HOI category form a single Gaussian cloud in CLIP's embedding space, so the estimated mean and variance faithfully summarize that category's visual patterns.

Editorial extensions

If this is right

HOI detectors that adopt InterProDa should improve disproportionately on rare and visually diverse categories, since the rare-split mAP gain (45.21 vs 42.67 full) is larger than the non-rare gain.
The approach transfers to other transformer-based HOI detectors: HOICLIP gains over 3 full mAP and ADA-CM gains over 3 full mAP with under 1-2 million extra parameters.
Zero-shot HOI detection also improves, with InterProDa outperforming prior prompt-based methods on both rare-first and non-rare-first unseen splits, even compared with multi-modal prompt methods.
Because the distribution representation is compatible with different space bases, the Gaussian choice is not critical: a Fourier basis yields similar performance, suggesting the benefit comes from the distributional query structure itself.

Reading between the lines

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

The paper does not test whether the Gaussian assumption is essential; if replaced with a mixture model or a non-parametric estimator, rare-category recall could improve further, since high-variance categories like 'hit a ball' appear visually multi-modal in the paper's own variance plots.
The paper's own ablation shows that replacing sampled queries with the distribution mean loses only about 0.28 mAP (42.39 vs 42.67), so the gains may come from variance-aware training rather than from stochastic sampling; a direct comparison of $\gamma=0$ versus larger $\gamma$ would settle that.
The three-way prompt structure (subject, object, interaction) is not confined to still images; the same distributional queries could be applied to video-based interaction detection or fine-grained action recognition, where intra-category variation is even larger.
The dynamic orthogonal loss only uses pooled category means; using full covariance structure or a learned distance between distributions could capture richer inter-category relationships.

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.

Desk Editor's Note

read the letter

You should know two things. The method genuinely moves the needle on HICO-DET, especially rare classes, and the ablation story is coherent. But Eq. (3) is sign-reversed: the 'orthogonal' loss actually drives cosine similarity up toward a fixed point, which is the opposite of the stated goal. That makes the central constraint non-reproducible as written.

The application of prompt distribution learning to HOI detection with subject/object/interaction prompt structures is new relative to ProDA and DreamDistribution. The gains on rare categories are large (45.21 rare mAP), and the portability results in Table 4 show real value: +2M params on HOICLIP and <1M on ADA-CM give meaningful boosts. The ablations are thorough and each component contributes. The learnable noise factor in reparameterization is a sensible fix for VAE-style instability.

The sign issue is not cosmetic: Table 3 credits Ldo with the largest single gain (39.56 to 41.85 Full). If the implementation matches the equation, the loss is attractive, not orthogonal. If it doesn't, the paper doesn't tell you. Either way, a reader cannot reproduce the result from the text. Also missing: standard deviations across runs, no code or weights, and the main Table 1 has formatting issues (numbers run together like '42.6745.21'). The Gaussian assumption for prompt embeddings is asserted with citations but not tested; that is a minor concern given the gains, but it is an assumption.

This paper deserves a serious referee despite my skepticism about the reproducibility. The core idea is worth evaluating and the empirical gains, if they hold, are valuable to the subfield. An accept should be conditioned on correcting Eq. (3) or explaining the actual loss, releasing code/weights, and adding variance estimates. I wouldn't cite it until the sign error is resolved.

Referee Report

2 major / 6 minor

Summary. The manuscript proposes InterProDa, a prompt-distribution-learning approach for HOI detection. For each HOI category, the method learns a collection of soft prompts fed through a CLIP text encoder, estimates a Gaussian distribution over the resulting embeddings, and samples from these distributions with a learnable noise factor to form category-distribution queries that are added to transformer decoder queries. Three structurally different prompt groups are used for subjects, objects, and interactions, and a dynamic orthogonal loss is applied to the object and interaction distribution spaces. Experiments report state-of-the-art results on HICO-DET and V-COCO, zero-shot results, ablations, and portability improvements when the module is added to HOICLIP and ADA-CM.

Significance. If the method is correctly implemented, the idea of distribution-valued queries is a plausible and useful step beyond fixed or learned point prompts, and the reported gains on rare HOI categories are substantial. The portability study in Table 4 is a practical strength, and the ablation set in Tables 3 and 5 is fairly comprehensive. However, the manuscript currently contains a sign-reversed loss equation for the component that Table 3 credits with the largest single improvement, and the reported benchmark numbers come from a single stochastic inference pass. Both issues must be resolved before the empirical claims can be relied upon.

major comments (2)

[§3.3, Eq. (3)] The dynamic orthogonal loss as written is sign-reversed with respect to the stated goal. Since Δ(i,j)=α(1−cos(i,j)), the penalty term becomes max(ε, α−(1+α)cos(i,j))². For α=0.5, a pair with cos=1 gives loss ε, while a pair with cos=0 gives 0.25, so well-separated pairs are penalized more than confused pairs; minimization drives cos toward approximately α/(1+α)=1/3. This is an attractive constraint, not an orthogonal one, and it contradicts the text's claim that Ldo 'increases the distance between ambiguous interaction categories'. This matters because Table 3 attributes the largest single ablation gain to +Ldo (39.56→41.85 Full). The authors must either correct Eq. (3), e.g. to max(ε, cos(i,j)−Δ(i,j))², or explicitly state how the implemented loss differs from the published equation; as written, the central constraint is not reproducible.
[§4.2–4.3] The reported results have no uncertainty quantification despite the inference procedure being stochastic. The sampling step S(P_c)=μ_c+γnσ_c draws fresh Gaussian noise at inference, and the implementation section specifies only a single random seed (42). The headline numbers in Table 1 (42.67 Full, 45.21 Rare) and Table 4 (41.71/42.92) are therefore single-rollout values. Since the claimed improvements over the closest baselines are on the order of 2–3 mAP, the paper should report mean and standard deviation over at least three seeds, or state that a fixed noise draw is used at test time; otherwise the SOTA claim is not statistically grounded.

minor comments (6)

[§3.3, Eq. (2)] The cosine similarity notation is ambiguous: the numerator is written as if it were a norm of the pair, and the denominator uses ℓ1 norms. Please write it as ⟨P̄_ci, P̄_cj⟩ / (‖P̄_ci‖_1‖P̄_cj‖_1+ε) and clarify whether P̄ is the mean-pooled token vector or a matrix.
[§3.3] The dynamic orthogonal loss is applied only to the object and interaction spaces, but no ablation shows the effect of excluding the subject space. A short experiment or a more detailed motivation would help.
[§3.2] The Gaussian assumption on prompt embeddings is asserted with citations to domain-generalization work, but no empirical check is provided for CLIP embeddings of HOI prompts. Given that Table 5(c) shows a Fourier basis achieves similar performance, the paper should either add a normality diagnostic or explicitly state that the method is robust to the distributional choice.
[Table 1, Table 2] Several table entries run numbers together, e.g. '42.6745.21' and '40.8839.58'; these need spacing or a column separator for readability.
[References] The reference list has formatting problems: 'Part et al. (2023)' in Table 1 should be 'Park et al. (2023)', 'Tamura et al.(2021)' is mis-formatted, and two different works are cited as 'Zhang et al. (2022)', making the in-text citations ambiguous.
[§5] The section heading contains a typo: 'Conlusion' should be 'Conclusion'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reported gains come from benchmark evaluation of an architectural/training method, not from fitted parameters or self-citation chains.

full rationale

InterProDa's derivation chain is not circular. The method proposes new trainable components—per-category soft prompt collections, Gaussian distribution estimation, a dynamic orthogonal loss, and a learnable-noise reparameterization—and validates them by measuring mAP/AP on held-out splits of HICO-DET and V-COCO against external methods. No equation in the paper reduces to an earlier fitted quantity; the Gaussian assumption in Section 3.2 is an explicit modeling choice supported by external citations to studies on neural feature geometry, and it is not used to define the benchmark target. The reported SOTA numbers are empirical outcomes of training and evaluation, so the 'prediction' is not statistically forced by a fitted input. The paper also does not rely on load-bearing self-citation: the cited prompt-distribution works (ProDA, DreamDistribution) are by different author groups, CLIP and the VAE reparameterization are standard external components, and no uniqueness theorem is invoked to forbid alternative designs. The dynamic orthogonal loss in Eq. (3) may be mathematically inconsistent with its stated goal of increasing inter-category separation, and the sign issue flagged by the skeptic is a legitimate correctness/reproducibility concern, but it is not a circularity: a flawed training objective is not an input that is later renamed as an output. Similarly, the Figure 2 observation that increasing pattern dimension helps is an empirical motivation, not a derivation of the final method. Overall, the central claim is self-contained against external benchmarks, so the circularity score is 0.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The method introduces no new physical or conceptual entities. It relies on standard neural network components and a distributional assumption for prompt embeddings. The free parameters are all hyperparameters or learnable scalars in the network, which are standard for a deep learning method.

free parameters (7)

gamma (learnable noise factor) = initialized at 1e-2, learned
Controls the balance between diversity and stability in the reparameterization sampling; value chosen by hand and then tuned.
alpha (dynamic margin tuning parameter) = 0.5
Set by hand in the dynamic orthogonal loss; no sensitivity analysis provided.
lambda (loss weight for Ldo) = 5e-2
Weight of the dynamic orthogonal loss; chosen by hand and tested in Table 5(d) for a few values.
K (number of prompts per category) = 8
Number of soft prompts per category; the paper claims it can be infinite, but it is set to 8 for efficiency. Ablation in Table 5(d) shows K=8 is best.
Ns (number of samples per distribution) = 2
Number of reparameterization samples; set to 2 and ablated in Table 5(d).
Nq (query dimension) = 80
Set to the number of object categories; no ablation is shown.
L (token length) = 77
Standard CLIP token length; chosen as a design decision.

assumptions (4)

domain assumption Gaussian distribution assumption for prompt embeddings
Section 3.2: 'we assume that for each HOI category, all the prompt embeddings describing this category follow the same Gaussian distribution.' No empirical justification is given.
domain assumption CLIP text embeddings with similar semantics remain close in feature space
Section 3.2 invokes this to justify the distributional representation, citing Lu et al. 2022.
domain assumption Increasing the query pattern dimension improves performance, motivating the distribution over patterns
Figure 2 shows this observation for a limited set of configurations; the paper extrapolates it to justify the distribution approach.
standard math Reparameterization with a learnable noise factor keeps the distribution differentiable
Standard reparameterization trick in Section 3.4, following Kingma and Welling 2022.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Orchestrating the Symphony of Prompt Distribution Learning for Human-Object Interaction Detection." pith.science (2026). https://pith.science/paper/3BEBGSJW

@misc{pith2026241208506,
  author       = {Pith},
  title        = {Pith review of: Orchestrating the Symphony of Prompt Distribution Learning for Human-Object Interaction Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3BEBGSJW}},
  note         = {Machine review of arXiv:2412.08506}
}

read the original abstract

Human-object interaction (HOI) detectors with popular query-transformer architecture have achieved promising performance. However, accurately identifying uncommon visual patterns and distinguishing between ambiguous HOIs continue to be difficult for them. We observe that these difficulties may arise from the limited capacity of traditional detector queries in representing diverse intra-category patterns and inter-category dependencies. To address this, we introduce the Interaction Prompt Distribution Learning (InterProDa) approach. InterProDa learns multiple sets of soft prompts and estimates category distributions from various prompts. It then incorporates HOI queries with category distributions, making them capable of representing near-infinite intra-category dynamics and universal cross-category relationships. Our InterProDa detector demonstrates competitive performance on HICO-DET and vcoco benchmarks. Additionally, our method can be integrated into most transformer-based HOI detectors, significantly enhancing their performance with minimal additional parameters.

Figures

Figures reproduced from arXiv: 2412.08506 by the authors.

**Figure 2.** Performance comparison w.r.t different settings of decoder queries on HICO-DET. We expand the query of our model with an additional intra-category pattern dimension. Each model is labeled with a format like 32 × 2, which indicates a model with a category (query) dimension of 32 and pattern dimension of 2. Models with the same color share identical parameters and overall query dimensions (for example, 32 × 2 equals… view at source ↗

**Figure 3.** The pipeline of InterProDa. We learn multiple groups of soft prompts for subject, object, and interaction categories. Then, we estimate the distributions of these category prompt embeddings and constrain them in a continuous feature space. Such an approach learns diverse intra-category patterns in each category distribution and captures the universal inter-category dependencies. We sampled from the learned distribut… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Visualization of intra-category variances of selected prompt distributions learned from HICO-DET. Each row refers to different categories, while each token refers to the variance of a single prompt. We select four distributions with the highest average variance and 4 …

**Figure 5.** Figure 5: T-SNE visualization of learned prompt distributions of 20 random interaction categories on HICODET. Each cluster with the same color refers to the category prompt embeddings belonging to a distinct HOI category distribution. The learned distribution space has clear m…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

EgoSim: Egocentric World Simulator for Embodied Interaction Generation
cs.CV 2026-04 conditional novelty 6.5 of 10

EgoSim generates spatially consistent egocentric interaction videos by conditioning a video diffusion model on updatable 3D point-cloud states and action keypoints extracted at scale from monocular videos.

Reference graph

Works this paper leans on

39 extracted references · 13 canonical work pages · cited by 1 Pith paper

[1]

Brown, T. B.; Mann, B.; Ryder, N.; Subbiah, M.; Kaplan, J.; Dhariwal, P.; Neelakantan, A.; Shyam, P.; Sastry, G.; Askell, A.; Agarwal, S.; Herbert-Voss, A.; Krueger, G.; Henighan, T.; Child, R.; Ramesh, A.; Ziegler, D. M.; Wu, J.; Winter, C.; Hesse, C.; Chen, M.; Sigler, E.; Litwin, M.; Gray, S.; Chess, B.; Clark, J.; Berner, C.; McCandlish, S.; Radford, ...

work page 2020
[2]

Cao, S.; Yin, Y.; Huang, L.; Liu, Y.; Zhao, X.; Zhao, D.; and Huang, K. 2023 a . Efficient-VQGAN: Towards High-Resolution Image Generation with Efficient Vision Transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 7368--7377

2023
[3]

Cao, Y.; Tang, Q.; Su, X.; Chen, S.; You, S.; Lu, X.; and Xu, C. 2023 b . Detecting Any Human-Object Interaction Relationship: Universal HOI Detector with Spatial Prompt Learning on Foundation Models. In Advances in Neural Information Processing Systems, volume 36, 739--751

2023
[4]

Chao, Y.-W.; Liu, Y.; Liu, X.; Zeng, H.; and Deng, J. 2018. Learning to Detect Human-Object Interactions. In 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), 381--389

2018
[5]

Gupta, S.; and Malik, J. 2015. Visual Semantic Role Labeling. arXiv:1505.04474

arXiv 2015
[6]

He, K.; Chen, X.; Xie, S.; Li, Y.; Dollár, P.; and Girshick, R. 2021. Masked Autoencoders Are Scalable Vision Learners. arXiv:2111.06377

arXiv 2021
[7]

Jia, C.; Yang, Y.; Xia, Y.; Chen, Y.-T.; Parekh, Z.; Pham, H.; Le, Q.; Sung, Y.-H.; Li, Z.; and Duerig, T. 2021. Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision. In Proceedings of the 38th International Conference on Machine Learning

work page 2021
[8]

Jiang, W.; Ren, W.; Tian, J.; Qu, L.; Wang, Z.; and Liu, H. 2024. Exploring Self- and Cross-Triplet Correlations for Human-Object Interaction Detection. arXiv:2401.05676

arXiv 2024

Show all 39 references

[9]

F.; Araki, J.; and Neubig, G

Jiang, Z.; Xu, F. F.; Araki, J.; and Neubig, G. 2020. How Can We Know What Language Models Know? Transactions of the Association for Computational Linguistics

2020
[10]

Kim, B.; Lee, J.; Kang, J.; Kim, E.; and Kim, H. J. 2021. HOTR: End-to-End Human-Object Interaction Detection With Transformers. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2021, virtual, June 19-25, 2021 , 74--83

2021
[11]

P.; and Welling, M

Kingma, D. P.; and Welling, M. 2022. Auto-Encoding Variational Bayes. arXiv:1312.6114

2022 arXiv
[12]

Lei, T.; Caba, F.; Chen, Q.; Jin, H.; Peng, Y.; and Liu, Y. 2023. Efficient Adaptive Human-Object Interaction Detection with Concept-guided Memory. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 6480--6490

2023
[13]

Lei, T.; Yin, S.; Peng, Y.; and Liu, Y. 2024. Exploring Conditional Multi-Modal Prompts for Zero-shot HOI Detection

2024
[14]

Li, J.; Li, D.; Xiong, C.; and Hoi, S. 2022. BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation. In ICML

2022
[15]

Li, L.; Wei, J.; Wang, W.; and Yang, Y. 2023. Neural-Logic Human-Object Interaction Detection. In Advances in Neural Information Processing Systems, volume 36, 21158--21171

2023
[16]

L.; and Liang, P

Li, X. L.; and Liang, P. 2021. Prefix-Tuning: Optimizing Continuous Prompts for Generation. arXiv:2101.00190

2021 arXiv
[17]

Liao, Y.; Zhang, A.; Lu, M.; Wang, Y.; Li, X.; and Liu, S. 2022. GEN-VLKT: Simplify Association and Enhance Interaction Understanding for HOI Detection. In IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022 , 20091--20100

2022
[18]

Liu, Y.; Xiong, Z.; Li, Y.; Lu, Y.; Tian, X.; and Zha, Z.-J. 2023 a . Category-Stitch Learning for Union Domain Generalization. ACM Trans. Multimedia Comput. Commun. Appl

2023
[19]

Liu, Y.; Xiong, Z.; Li, Y.; Tian, X.; and Zha, Z.-J. 2023 b . Domain Generalization Via Encoding and Resampling in a Unified Latent Space. IEEE Transactions on Multimedia, 25: 126--139

2023
[20]

Lu, Y.; Liu, J.; Zhang, Y.; Liu, Y.; and Tian, X. 2022. Prompt Distribution Learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 5206--5215

2022
[21]

Ning, S.; Qiu, L.; Liu, Y.; and He, X. 2023. HOICLIP: Efficient Knowledge Transfer for HOI Detection with Vision-Language Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,bCVPR 2023, 23507--23517

2023
[22]

K.; and Kim, H

Park, J.; Lee, S.; Heo, H.; Choi, H. K.; and Kim, H. J. 2022. Consistency Learning via Decoding Path Augmentation for Transformers in Human Object Interaction Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022

2022
[23]

(2023) Park, J.; Park, J.-W.; and Lee, J.-S

Part et al. (2023) Park, J.; Park, J.-W.; and Lee, J.-S. 2023. ViPLO: Vision Transformer Based Pose-Conditioned Self-Loop Graph for Human-Object Interaction Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 17152--17162

2023
[24]

W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; Krueger, G.; and Sutskever, I

Radford, A.; Kim, J. W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; Krueger, G.; and Sutskever, I. 2021. Learning Transferable Visual Models From Natural Language Supervision. In Proceedings of the 38th International Conferen...

2021
[25]

Rombach, R.; Blattmann, A.; Lorenz, D.; Esser, P.; and Ommer, B. 2022. High-Resolution Image Synthesis with Latent Diffusion Models. arXiv:2112.10752

2022 arXiv
[26]

L.; Wallace, E.; and Singh, S

Shin, T.; Razeghi, Y.; Logan IV, R. L.; Wallace, E.; and Singh, S. 2020. A uto P rompt: E liciting K nowledge from L anguage M odels with A utomatically G enerated P rompts. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

2020
[27]

Tamura et al.(2021) Masato, T.; Hiroki, O.; and Tomoaki, Y. 2021. QPIC: Query-Based Pairwise Human-Object Interaction Detection With Image-Wide Contextual Information. In IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2021, virtual, June 19-25, 2021 , 10410--10419

2021
[28]

Wang, G.; Guo, Y.; Xu, Z.; and Kankanhalli, M. 2024. Bilateral Adaptation for Human-Object Interaction Detection with Occlusion-Robustness. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 27970--27980

2024
[29]

Williams, C.; and Barber, D. 1998. Bayesian classification with Gaussian processes. IEEE Transactions on Pattern Analysis and Machine Intelligence, 20(12): 1342--1351

1998
[30]

Xie, C.; Zeng, F.; Hu, Y.; Liang, S.; and Wei, Y. 2023. Category Query Learning for Human-Object Interaction Classification. In CVPR

2023
[31]

Yang, J.; Li, B.; Zeng, A.; Zhang, L.; and Zhang, R. 2024. Open-World Human-Object Interaction Detection via Multi-modal Prompts. arXiv preprint arXiv:2406.07221

2024 arXiv
[32]

Yuan, H.; Zhang, S.; Wang, X.; Albanie, S.; Pan, Y.; Feng, T.; Jiang, J.; Ni, D.; Zhang, Y.; and Zhao, D. 2023. RLIPv2: Fast Scaling of Relational Language-Image Pre-Training. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 21649--21661

2023
[33]

Z.; Yuan, Y.; Campbell, D.; Zhong, Z.; and Gould, S

Zhang, F. Z.; Yuan, Y.; Campbell, D.; Zhong, Z.; and Gould, S. 2023. Exploring Predicate Visual Context in Detecting of Human-Object Interactions. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 10411--10421

2023
[34]

Zhang, Y.; Gong, M.; Liu, T.; Niu, G.; Tian, X.; Han, B.; Scholkopf, B.; and Zhang, K. 2022. Adversarial Robustness through the Lens of Causality. ArXiv, abs/2106.06196

2022 arXiv
[35]

(2022) Zhang, F

Zhang et al. (2022) Zhang, F. Z.; Campbell, D.; and Gould, S. 2022. Efficient Two-Stage Detection of Human-Object Interactions with a Novel Unary-Pairwise Transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 20104--20112

2022
[36]

N.; Xiao, Y.; Xu, J.; Jiang, X.; Yang, Y.; Li, D.; Itti, L.; Vineet, V.; and Ge, Y

Zhao, B. N.; Xiao, Y.; Xu, J.; Jiang, X.; Yang, Y.; Li, D.; Itti, L.; Vineet, V.; and Ge, Y. 2023. DreamDistribution: Prompt Distribution Learning for Text-to-Image Diffusion Models. arXiv:2312.14216

2023 arXiv
[37]

C.; and Liu, Z

Zhou, K.; Yang, J.; Loy, C. C.; and Liu, Z. 2022. Learning to Prompt for Vision-Language Models. International Journal of Computer Vision (IJCV)

2022
[38]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...
[39]

write newline

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

Pith tools

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

[1] [1]

Brown, T. B.; Mann, B.; Ryder, N.; Subbiah, M.; Kaplan, J.; Dhariwal, P.; Neelakantan, A.; Shyam, P.; Sastry, G.; Askell, A.; Agarwal, S.; Herbert-Voss, A.; Krueger, G.; Henighan, T.; Child, R.; Ramesh, A.; Ziegler, D. M.; Wu, J.; Winter, C.; Hesse, C.; Chen, M.; Sigler, E.; Litwin, M.; Gray, S.; Chess, B.; Clark, J.; Berner, C.; McCandlish, S.; Radford, ...

work page 2020

[2] [2]

Cao, S.; Yin, Y.; Huang, L.; Liu, Y.; Zhao, X.; Zhao, D.; and Huang, K. 2023 a . Efficient-VQGAN: Towards High-Resolution Image Generation with Efficient Vision Transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 7368--7377

2023

[3] [3]

Cao, Y.; Tang, Q.; Su, X.; Chen, S.; You, S.; Lu, X.; and Xu, C. 2023 b . Detecting Any Human-Object Interaction Relationship: Universal HOI Detector with Spatial Prompt Learning on Foundation Models. In Advances in Neural Information Processing Systems, volume 36, 739--751

2023

[4] [4]

Chao, Y.-W.; Liu, Y.; Liu, X.; Zeng, H.; and Deng, J. 2018. Learning to Detect Human-Object Interactions. In 2018 IEEE Winter Conference on Applications of Computer Vision (WACV), 381--389

2018

[5] [5]

Gupta, S.; and Malik, J. 2015. Visual Semantic Role Labeling. arXiv:1505.04474

arXiv 2015

[6] [6]

He, K.; Chen, X.; Xie, S.; Li, Y.; Dollár, P.; and Girshick, R. 2021. Masked Autoencoders Are Scalable Vision Learners. arXiv:2111.06377

arXiv 2021

[7] [7]

Jia, C.; Yang, Y.; Xia, Y.; Chen, Y.-T.; Parekh, Z.; Pham, H.; Le, Q.; Sung, Y.-H.; Li, Z.; and Duerig, T. 2021. Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision. In Proceedings of the 38th International Conference on Machine Learning

work page 2021

[8] [8]

Jiang, W.; Ren, W.; Tian, J.; Qu, L.; Wang, Z.; and Liu, H. 2024. Exploring Self- and Cross-Triplet Correlations for Human-Object Interaction Detection. arXiv:2401.05676

arXiv 2024