Pith. sign in

REVIEW 5 major objections 9 minor 43 references

Exploring Transferable Homogeneous Groups for Compositional Zero-Shot Learning

T0 review · 5 major / 9 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Grouping states and objects into homogeneous sub-groups, each with its own learned representation center, balances transferability and discriminability in compositional zero-shot learning and sets new state-of-the-art scores on three…

desk verdict HGRL is a credible middle-ground CZSL model with consistently strong benchmark numbers; the main weakness is that its GloVe-based grouping proxy is under-validated and the results are single-run. read the letter →

arxiv 2501.10695 v2 pith:7DQBTJHS submitted 2025-01-18 cs.CV

Zhijie Rao , Jingcai Guo , Miaoge Li , Yang Chen This is my paper · ORCID
classification cs.CV
keywords CompositionalZero-ShotLearningHomogeneousGroupRepresentationMixture-of-ExpertsVision-LanguageModelsPromptGloVewordembeddingsTransferabilityandDiscriminabilityDomainAdaptation
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 the two dominant paradigms for compositional zero-shot learning—one shared representation per state or object ('all-to-one') versus a separate representation per state-object pair ('one-to-one')—each sacrifice either transferability or discriminability. It proposes Homogeneous Group Representation Learning (HGRL), which clusters states and objects into a small number of homogeneous sub-groups, learns a distributed representation center per group, and adaptively aggregates features through a mixture-of-experts router guided by word-embedding similarities. On the MIT-States, UT-Zappos, and C-GQA benchmarks, HGRL reports higher area-under-the-curve (AUC) and harmonic-mean (HM) accuracy than prior CLIP-based methods, with the largest closed-world gain on UT-Zappos (46.8 AUC versus 44.5 for the previous best). A sympathetic reader would care because the result suggests that the transferability–discriminability tradeoff in CZSL is not a zero-sum game: a few learned group centers can carry both shared and group-specific semantics.

What carries the argument

The load-bearing mechanism is the adaptive grouping loop: a text co-occurrence probability graph (TCPG) derived from GloVe cosine similarity among category names, thresholded at 0.5, defines an adjacency matrix over a batch of samples (Eq. 4); the route network in the Group-Aware Visual Representation (GAVR) module assigns each sample to the K most confident experts, and group-specific features are a weighted sum of those experts' outputs fused with the original feature (Eqs. 1–3). The same route confidences are reused in Group-Aware Pair Enhancement (GAPE) to gate which samples may augment each other (Eq. 14), and the Decoupled Group Prompt (DGP) appends a learnable per-group prompt token so each group has its own text prototype rather than one shared prototype per category. The theoretical section frames the design as minimizing a domain-adaptation upper bound by increasing the probability that an unseen domain's homogeneous group matches the seen group that transfers knowledge to it.

What would settle it

One decisive experiment: replace the GloVe adjacency in Eq. (4) with an adjacency computed from visual similarity of held-out CLIP image features (or from human-annotated superclass labels) and retrain HGRL; if the visual- or label-guided version does not match or beat the GloVe version, then word-embedding similarity is not doing the work the paper claims. Alternatively, one can inspect the discovered groups for a dataset such as MIT-States and measure intra-group visual similarity of the learned representations: if the groups are not internally visually consistent, the grouping assumption fails.

Watch

Extended reading notes

Core claim

The central claim is that state and object representations in compositional zero-shot learning should be learned as multiple homogeneous sub-group representations rather than as a single global prototype or as per-instance dependencies. A route network assigns each sample to a small set of experts, and a text co-occurrence probability graph built from GloVe embeddings of category names provides soft supervision for which categories belong together by aggregating features of compatible samples. Each group receives its own learnable prompt context (Decoupled Group Prompt), and Group-Aware Pair Enhancement mixes state and object features only for pairs whose group-confidence vectors are compatible. The paper reports state-of-the-art closed-world AUC of 23.3 on MIT-States, 46.8 on UT-Zappos, and 16.3 on C-GQA, with corresponding harmonic means of 40.2, 59.0, and 34.6, and open-world AUC gains of up to 4.2 points over the previous best on UT-Zappos.

Load-bearing premise

The method assumes that the cosine similarity between GloVe word embeddings of two category names (for instance 'dog' and 'tiger') is a reliable guide to whether images of those categories in the same state look alike, so that aggregating visually compatible samples across groups transfers knowledge correctly.

Editorial extensions

If this is right

  • The transferability–discriminability tradeoff in CZSL is not a zero-sum game: a small number of learned group centers capture both shared and group-specific semantics, which is why HGRL beats both the all-to-one and one-to-one baselines.
  • Group-level prompts provide a supervision-free way to inject structured prior knowledge into vision-language models, since the grouping signal comes from word co-occurrence statistics rather than hierarchical labels.
  • The adjacency construction (GloVe similarity thresholded at 0.5) is reusable across datasets and requires no extra annotation, making the method easy to adapt to other attribute-object tasks.
  • The largest reported gains appear on UT-Zappos, a small dataset with strong visual grouping structure, suggesting the benefit scales with how cleanly categories cluster in feature space.
  • Fusing backbone, pair, and state × object branch scores at inference (Eq. 22) yields consistent improvements over any single branch, so the components are complementary.

Reading between the lines

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

  • If the paper is right, replacing the GloVe-based adjacency with a vision-derived grouping (e.g., clustering CLIP visual features of seen images) could make the method work on datasets where word co-occurrence and perceptual similarity diverge; the paper does not run that comparison.
  • The theory suggests a principled way to choose the number of groups: stop increasing ks and ko when validation AUC saturates, rather than fixing them at 3 or 5; the sensitivity plots in Fig. 3 hint at this but do not test it.
  • The compatibility gate in GAPE could be applied symmetrically to object-state pairs in the other direction or extended to multi-hop neighbors in the adjacency graph, which the paper does not explore.
  • Because the method builds on frozen CLIP features plus GloVe, it is a drop-in module for other CZSL baselines; one could measure the marginal gain of adding HGRL to a non-CLIP method, which the paper does not do.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 9 minor

Summary. The paper proposes Homogeneous Group Representation Learning (HGRL) for compositional zero-shot learning, motivated by the claim that existing all-to-one and one-to-one representation paradigms create a transferability-discriminability trade-off. HGRL has three components: Group-Aware Visual Representation (GA VR), which uses a mixture-of-experts network with a GloVe-based text co-occurrence graph to form homogeneous groups and aggregate visual features; Decoupled Group Prompt (DGP), which learns a separate prompt context for each group; and Group-Aware Pair Enhancement (GAPE), which augments state and object features using group-compatibility-weighted pairs. The method is evaluated in closed- and open-world settings on MIT-States, UT-Zappos, and C-GQA with a CLIP ViT-L/14 backbone, reporting state-of-the-art AUC and HM numbers and ablations over the three components.

Significance. If the reported results hold, HGRL is a useful contribution to CZSL: the grouping perspective is clearly different from the all-to-one/one-to-one dichotomy, the three modules are coherent and each is supported by the Table 2 ablations, and the gains on UT-Zappos and C-GQA are large enough to be meaningful even without formal tests. The paper also provides a fair comparison with recent CLIP-based methods on three standard benchmarks and includes visualization analyses. However, the absence of repeated-run statistics, the undocumented hyperparameter selection, the unvalidated GloVe-based grouping proxy, and the loose theoretical section mean the central claim of state-of-the-art performance is not yet fully established.

major comments (5)
  1. [Table 1 and Section 4.2] All results in Table 1 and Table 2 appear to be single-run point estimates: no standard deviations, seeds, or significance tests are reported, and the text does not state how many runs each number is averaged over. This is load-bearing because the closed-world MIT-States improvement over the second-best method is only 0.8 AUC (23.3 vs 22.5) and 0.9 HM (40.2 vs 39.3), a margin that can easily be within run-to-run noise for CLIP-based training. Please report mean and standard deviation over multiple seeds, or otherwise provide significance evidence for the claimed state-of-the-art results.
  2. [Section 4.1 and Eq. (2)] Hyperparameter selection is not reproducible as described. In Section 4.1, λ is set to 1.0 for UT-Zappos and 0.1 for MIT-States and C-GQA, while ks and ko are 3 for UT-Zappos and 5 for the other datasets, and no validation split or selection criterion is given; additionally, the top-K expert filtering used in Eq. (2) is never specified. Because the test sets include unseen compositions, tuning these values on the test set would inflate the reported numbers. The paper should state the validation protocol and report results for the chosen values, ideally with sensitivity curves that include error bars.
  3. [Section 3.2, Eq. (4)] Eq. (4) constructs the entire grouping mechanism from GloVe cosine similarity: A_ij = M^o_ij * I[M^o_ij >= 0.5] * I[s_i = s_j], with M^o_ij = cos(v_i^o, v_j^o). The paper asserts that word co-occurrence probability 'largely matches' membership in the same perceptual superclass, but the only empirical support is the w/o TCPG ablation in Table 2, which shows only that removing the graph hurts performance. It does not show that the GloVe groups are visually coherent; a generic feature-aggregation regularizer could produce a similar drop. Please validate the proxy directly (for example, by comparing GloVe-based group adjacency with clusters in the visual feature space or with group annotations), and also state how GloVe embeddings are computed for multiword UT-Zappos labels such as 'Shoes.Sneakers.and.Athletic.Shoes'.
  4. [Section 3.6, Eq. (24)] Section 3.6 is not a rigorous derivation. The notation d(S,U | p(...)) is not defined, the claimed decomposition of d(S,U) into conditional divergences is asserted rather than proved, and the conclusion d(S*,U) <= d(S,U) does not follow without assumptions on the domains and the group-conditional measures. Since this section is explicitly titled 'Theoretical Insights,' it should either be made into a proper theorem with stated assumptions and proof or be recast as motivation with the 'theoretical' claim softened.
  5. [Section 3.5, Eq. (21)] Eq. (21), the inference formula for pstate(s|xs), is inconsistent with the training loss in Eq. (12): the numerator in Eq. (12) is weighted by the route confidence p^{Sj}_r, but the numerator in Eq. (21) is unweighted and contains a stray dot. The object-branch inference likely has the same issue. Please correct the formulas and verify that the inference used for the reported numbers matches the trained objective.
minor comments (9)
  1. [Abstract and Introduction] The abstract and introduction contain several grammatical errors, such as 'Conditional dependency present one of the trickiest problems' and 'a new perspective formulates state (object) representation learning'; these should be copy-edited.
  2. [Section 3.2, Fig. 1, Table 1, Section 4.2] There are several typos: 'memebers' in Section 3.2, 'Visualizaiton' in the Fig. 1 caption, 'there datasets' in the Table 1 caption, and 'CGQ' instead of C-GQA in Section 4.2.
  3. [Eq. (5)] Eq. (5) calls Λ a 'unit diagonal matrix'; this should be the identity matrix.
  4. [Eq. (2)] In Eq. (2), the sum over the selected K experts is written without an index set; please make explicit that the sum runs over the top-K experts selected by the route scores.
  5. [Section 3.2] The symmetric object-branch analogue of Eq. (4), which should use M^s_ij and I[o_i = o_j], is not written out; please state it explicitly.
  6. [Fig. 3] Fig. 3(a)-(b) report 'state accuracy' and 'object accuracy' without defining these metrics, and the λ-sensitivity plots in Fig. 3(c)-(d) have no error bars and no legend indicating which curve is which.
  7. [Section 3.2, Eq. (4)] The term 'text co-occurrence probability graph' is misleading because Eq. (4) uses a cosine similarity threshold, not an estimated co-occurrence probability; align the terminology with the actual computation.
  8. [References] ProCC [Huo et al., 2024] appears in the reference list but is never cited in the main text.
  9. [Table 2] The ablation study in Table 2 omits MIT-States; since the main claim covers all three datasets, it would be informative to include MIT-States ablations as well.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the benchmark results are measured, the GloVe group prior is an external assumption rather than a fitted output, and the only self-citations are non-load-bearing baselines/related work.

full rationale

HGRL's central claim is empirical: the reported AUC/HM numbers are measured on fixed test splits, not produced by a fitted parameter later renamed as a prediction. The group adjacency in Eq. (4), A_ij = M^o_ij * I[M^o_ij >= 0.5] * I[s_i = s_j], is built from GloVe cosine similarities, which are fixed word embeddings external to the training objective; the model then learns route confidences (Eq. 1), expert features (Eq. 2), and group prompts (Eqs. 9-10) under standard InfoNCE losses (Eqs. 12, 13, 17). The inference rule (Eq. 22) combines backbone, pair, and state/object branch scores, and no term in that rule is defined as the target value being predicted. The domain-adaptation discussion in Sec. 3.6 is a post hoc interpretation of an upper bound and is not used to derive the method, so it does not create a derivation loop. The only self-citations are to DFSP and ProCC as comparison/related methods; these are not load-bearing for HGRL's design or results. The unvalidated GloVe-as-visual-group-proxy assumption is a genuine robustness concern, but it is an external assumption rather than a circular reduction: the paper never defines GloVe similarity in terms of the learned groups or the target compositions. No circular step can be exhibited by quoting the paper, and the empirical claims are self-contained against the benchmarks reported.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claim rests on several unverified modeling choices: the GloVe word co-occurrence proxy for semantic grouping, hand-picked expert counts, a fixed similarity threshold, and a post hoc domain-adaptation argument that is not used to derive the objective. These are not fatal, but they shift the burden of proof from a derivation to benchmark performance.

free parameters (4)
  • lambda (loss weight) = 1.0 (UT-Zappos), 0.1 (MIT-States, C-GQA)
    Loss weight in Eq. (19) chosen per dataset to maximize reported HM/AUC; no validation split described.
  • ks, ko (group counts) = 3 (UT-Zappos), 5 (MIT-States, C-GQA)
    Number of expert groups per state/object branch, hand-set in §4.1 and explored only after the fact in Fig. 3.
  • zeta (similarity threshold) = 0.5
    Threshold in Eqs. (4) and (14) fixed by hand to prune group edges; no sensitivity analysis is provided.
  • K (top experts) = not specified
    Eq. (2) selects K experts but implementation details never state K, so replication is ambiguous.
assumptions (4)
  • domain assumption Cosine similarity of GloVe word embeddings approximates the probability that two object or state names belong to the same real-world superclass.
    Introduced in §3.2 ('the probability of co-occurrence between words largely matches the probability that they belong to the same superclass'); it is the basis for Eq. (4) and is not validated directly.
  • ad hoc to paper A small fixed number of experts (3 or 5) can cover the latent homogeneous groups for states and objects.
    The paper states that the true group count is agnostic (§4.3) and sets ks, ko by hand; no criterion ensures the chosen count covers the data.
  • standard math The Ben-David domain adaptation bound (Eq. 23) applies to CZSL state and object shifts.
    The bound is taken as given from Ben-David et al. 2006 and is not re-derived for the CZSL setting.
  • ad hoc to paper The divergence decomposition in Eq. (24), separating same-group and different-group terms, holds and is meaningful for CZSL.
    The decomposition is asserted without proof and is not connected to the training loss; it is a post hoc framing rather than a derivation.
invented entities (1)
  • latent homogeneous groups
    purpose: Latent clusters of states and objects that share perceptual properties; used to learn group-specific visual features and prompts.
    The groups are inferred by a learned router plus GloVe similarities and are never compared against an external taxonomy. Evidence is indirect (ablation and t-SNE on two UT-Zappos categories), and no external falsifiable prediction is made about group membership.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Exploring Transferable Homogeneous Groups for Compositional Zero-Shot Learning." pith.science (2026). https://pith.science/paper/7DQBTJHS

@misc{pith2026250110695,
  author       = {Pith},
  title        = {Pith review of: Exploring Transferable Homogeneous Groups for Compositional Zero-Shot Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7DQBTJHS}},
  note         = {Machine review of arXiv:2501.10695}
}
read the original abstract

Conditional dependency present one of the trickiest problems in Compositional Zero-Shot Learning, leading to significant property variations of the same state (object) across different objects (states). To address this problem, existing approaches often adopt either all-to-one or one-to-one representation paradigms. However, these extremes create an imbalance in the seesaw between transferability and discriminability, favoring one at the expense of the other. Comparatively, humans are adept at analogizing and reasoning in a hierarchical clustering manner, intuitively grouping categories with similar properties to form cohesive concepts. Motivated by this, we propose Homogeneous Group Representation Learning (HGRL), a new perspective formulates state (object) representation learning as multiple homogeneous sub-group representation learning. HGRL seeks to achieve a balance between semantic transferability and discriminability by adaptively discovering and aggregating categories with shared properties, learning distributed group centers that retain group-specific discriminative features. Our method integrates three core components designed to simultaneously enhance both the visual and prompt representation capabilities of the model. Extensive experiments on three benchmark datasets validate the effectiveness of our method.

Figures

Figures reproduced from arXiv: 2501.10695 by the authors.

Figure 1
Figure 1. (a) Motivation. Humans utilize higher-order knowledge structure to perform hierarchical clustering, enabling effective analogies and reasoning. (b) Visualization of the state—old. The semantic structure of homogeneous groups is well maintained in deep feature space, e.g., animals are naturally clustered together. (c) Visualizaiton of the object—apple. Similarly, apples in various states cluster in groups due to visu… view at source ↗
Figure 2
Figure 2. Overview of the proposed method, which comprises three main components to enhance both visual and prompt representations. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. (a) The effect of group number for state branch. (b) The effect of group number for object branch. (c) The sensitivity of [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: (a) Attention visualization for GAVR. (b-c) T-SNE analysis for DGP. Red pentagrams indicate group prompt representations. Dots [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 36 canonical work pages

  1. [1]

    A causal view of compositional zero-shot recognition

    [Atzmon et al., 2020] Yuval Atzmon, Felix Kreuk, Uri Shalit, and Gal Chechik. A causal view of compositional zero-shot recognition. Advances in Neural Information Processing Systems, 33:1462–1473,

    work page 2020

  2. [5]

    Inferring analogous attributes

    [Chen and Grauman, 2014] Chao-Yeh Chen and Kristen Grauman. Inferring analogous attributes. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 200–207,

    work page 2014

  3. [13]

    Discovering states and transformations in im- age collections

    [Isola et al., 2015] Phillip Isola, Joseph J Lim, and Edward H Adelson. Discovering states and transformations in im- age collections. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 1383– 1391,

    work page 2015

  4. [16]

    Hierarchical visual primitive experts for compositional zero-shot learning

    [Kim et al., 2023] Hanjae Kim, Jiyoung Lee, Seongheon Park, and Kwanghoon Sohn. Hierarchical visual primitive experts for compositional zero-shot learning. In Proceed- ings of the IEEE/CVF International Conference on Com- puter Vision, pages 5675–5685,

    work page 2023

  5. [17]

    Symmetry and group in attribute-object com- positions

    [Li et al., 2020] Yong-Lu Li, Yue Xu, Xiaohan Mao, and Cewu Lu. Symmetry and group in attribute-object com- positions. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11316– 11325,

    work page 2020

  6. [18]

    Distilled reverse attention network for open-world compositional zero-shot learning

    [Li et al., 2023] Yun Li, Zhe Liu, Saurav Jha, and Lina Yao. Distilled reverse attention network for open-world compositional zero-shot learning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1782–1791,

    work page 2023

  7. [19]

    Context-based and diversity-driven specificity in composi- tional zero-shot learning

    [Li et al., 2024] Yun Li, Zhe Liu, Hang Chen, and Lina Yao. Context-based and diversity-driven specificity in composi- tional zero-shot learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17037–17046,

    work page 2024

  8. [20]

    Simple primitives with feasibility-and contextuality- dependence for open-world compositional zero-shot learn- ing

    [Liu et al., 2023] Zhe Liu, Yun Li, Lina Yao, Xiaojun Chang, Wei Fang, Xiaojun Wu, and Abdulmotaleb El Sad- dik. Simple primitives with feasibility-and contextuality- dependence for open-world compositional zero-shot learn- ing. IEEE Transactions on Pattern Analysis and Machine Intelligence,

    work page 2023

Show all 43 references
  1. [21]

    Decomposed soft prompt guided fusion en- hancing for compositional zero-shot learning

    [Lu et al., 2023] Xiaocheng Lu, Song Guo, Ziming Liu, and Jingcai Guo. Decomposed soft prompt guided fusion en- hancing for compositional zero-shot learning. In Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 23560–23569,

    2023

  2. [22]

    From red wine to red tomato: Composition with context

    [Misra et al., 2017] Ishan Misra, Abhinav Gupta, and Mar- tial Hebert. From red wine to red tomato: Composition with context. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 1792– 1801,

    2017

  3. [24]

    Attributes as operators: factorizing unseen attribute-object compositions

    [Nagarajan and Grauman, 2018] Tushar Nagarajan and Kris- ten Grauman. Attributes as operators: factorizing unseen attribute-object compositions. In Proceedings of the Euro- pean Conference on Computer Vision (ECCV), pages 169– 185,

    2018

  4. [26]

    Learning to compose soft prompts for compositional zero-shot learning

    [Nayak et al., 2022] Nihal V Nayak, Peilin Yu, and Stephen H Bach. Learning to compose soft prompts for compositional zero-shot learning. arXiv preprint arXiv:2204.03574,

    2022 arXiv

  5. [27]

    Chils: Zero- shot image classification with hierarchical label sets

    [Novack et al., 2023] Zachary Novack, Julian McAuley, Zachary Chase Lipton, and Saurabh Garg. Chils: Zero- shot image classification with hierarchical label sets. In International Conference on Machine Learning , pages 26342–26362. PMLR,

    2023

  6. [28]

    A hierarchical classification ant colony algorithm for predicting gene ontology terms

    [Otero et al., 2009] Fernando EB Otero, Alex A Freitas, and Colin G Johnson. A hierarchical classification ant colony algorithm for predicting gene ontology terms. In Evolu- tionary Computation, Machine Learning and Data Min- ing in Bioinformatics: 7th European Conference, Evo...

    2009

  7. [30]

    Glove: Global vec- tors for word representation

    [Pennington et al., 2014] Jeffrey Pennington, Richard Socher, and Christopher D Manning. Glove: Global vec- tors for word representation. In Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP), pages 1532–1543,

    2014

  8. [31]

    Learning transferable visual models from nat- ural language supervision

    [Radford et al., 2021] Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agar- wal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from nat- ural language supervision. In International confere...

    2021

  9. [32]

    Independent prototype propagation for zero- shot compositionality

    [Ruis et al., 2021] Frank Ruis, Gertjan Burghouts, and Doina Bucur. Independent prototype propagation for zero- shot compositionality. Advances in Neural Information Processing Systems, 34:10641–10653,

    2021

  10. [33]

    Disentangling visual embeddings for at- tributes and objects

    [Saini et al., 2022] Nirat Saini, Khoi Pham, and Abhinav Shrivastava. Disentangling visual embeddings for at- tributes and objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13658–13667,

    2022

  11. [34]

    The graph neural network model

    [Scarselli et al., 2008] Franco Scarselli, Marco Gori, Ah Chung Tsoi, Markus Hagenbuchner, and Gabriele Monfardini. The graph neural network model. IEEE transactions on neural networks, 20(1):61–80,

    2008

  12. [37]

    Adversarial fine-grained com- position learning for unseen attribute-object recognition

    [Wei et al., 2019] Kun Wei, Muli Yang, Hao Wang, Cheng Deng, and Xianglong Liu. Adversarial fine-grained com- position learning for unseen attribute-object recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 3741–3749,

    2019

  13. [38]

    Prompting large pre-trained vision-language models for compositional concept learning

    [Xu et al., 2022] Guangyue Xu, Parisa Kordjamshidi, and Joyce Chai. Prompting large pre-trained vision-language models for compositional concept learning. arXiv preprint arXiv:2211.05077,

    2022 arXiv

  14. [39]

    Detclipv3: Towards versatile generative open- vocabulary object detection

    [Yao et al., 2024] Lewei Yao, Renjie Pi, Jianhua Han, Xi- aodan Liang, Hang Xu, Wei Zhang, Zhenguo Li, and Dan Xu. Detclipv3: Towards versatile generative open- vocabulary object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, p...

    2024

  15. [40]

    Fine-grained visual comparisons with local learning

    [Yu and Grauman, 2014] Aron Yu and Kristen Grauman. Fine-grained visual comparisons with local learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 192–199,

    2014

  16. [41]

    Learning in- variant visual representations for compositional zero-shot learning

    [Zhang et al., 2022] Tian Zhang, Kongming Liang, Ruoyi Du, Xian Sun, Zhanyu Ma, and Jun Guo. Learning in- variant visual representations for compositional zero-shot learning. In European Conference on Computer Vision , pages 339–355. Springer,

    2022

  17. [42]

    Learning to prompt for vision-language models

    [Zhou et al., 2022] Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. Learning to prompt for vision-language models. International Journal of Computer Vision, 130(9):2337–2348,

    2022

  18. [43]

    B-cnn: branch convolutional neural network for hierarchical clas- sification

    [Zhu and Bain, 2017] Xinqi Zhu and Michael Bain. B-cnn: branch convolutional neural network for hierarchical clas- sification. arXiv preprint arXiv:1709.09890, 2017

    2017 arXiv

  19. [1991]

    Retrieval-augmented primitive representa- tions for compositional zero-shot learning

    [Jing et al., 2024] Chenchen Jing, Yukun Li, Hao Chen, and Chunhua Shen. Retrieval-augmented primitive representa- tions for compositional zero-shot learning. In Proceedings of the AAAI Conference on Artificial Intelligence,

    2024

  20. [2004]

    Dual part discovery network for zero-shot learning

    [Ge et al., 2022] Jiannan Ge, Hongtao Xie, Shaobo Min, Pandeng Li, and Yongdong Zhang. Dual part discovery network for zero-shot learning. In Proceedings of the 30th ACM International Conference on Multimedia, pages 3244–3252,

    2022

  21. [2006]

    An empirical study on large-scale multi-label text classification including few and zero-shot labels

    [Chalkidis et al., 2020] Ilias Chalkidis, Manos Fergadiotis, Sotiris Kotitsas, Prodromos Malakasiotis, Nikolaos Ale- tras, and Ion Androutsopoulos. An empirical study on large-scale multi-label text classification including few and zero-shot labels. In Proceedings of the 2020 ...

    2020

  22. [2008]

    A survey of hierarchical classification across different ap- plication domains

    [Silla and Freitas, 2011] Carlos N Silla and Alex A Freitas. A survey of hierarchical classification across different ap- plication domains. Data mining and knowledge discovery, 22:31–72,

    2011

  23. [2009]

    Compositional zero-shot learning using multi-branch graph convolution and cross-layer knowledge sharing

    [Panda and Mukherjee, 2024] Aditya Panda and Dipti Prasad Mukherjee. Compositional zero-shot learning using multi-branch graph convolution and cross-layer knowledge sharing. Pattern Recognition , 145:109916,

    2024

  24. [2011]

    Learning conditional attributes for compositional zero-shot learning

    [Wang et al., 2023] Qingsheng Wang, Lingqiao Liu, Chenchen Jing, Hao Chen, Guoqiang Liang, Peng Wang, and Chunhua Shen. Learning conditional attributes for compositional zero-shot learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, p...

    2023

  25. [2014]

    Transferability vs

    [Chen et al., 2019] Xinyang Chen, Sinan Wang, Mingsheng Long, and Jianmin Wang. Transferability vs. discrim- inability: Batch spectral penalization for adversarial do- main adaptation. In International conference on machine learning, pages 1081–1090. PMLR,

    2019

  26. [2015]

    Adaptive mix- tures of local experts

    [Jacobs et al., 1991] Robert A Jacobs, Michael I Jordan, Steven J Nowlan, and Geoffrey E Hinton. Adaptive mix- tures of local experts. Neural computation, 3(1):79–87,

    1991

  27. [2017]

    Learning graph embeddings for compositional zero-shot learning

    [Naeem et al., 2021] Muhammad Ferjad Naeem, Yongqin Xian, Federico Tombari, and Zeynep Akata. Learning graph embeddings for compositional zero-shot learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 953–962,

    2021

  28. [2018]

    Recognizing unseen attribute-object pair with generative model

    [Nan et al., 2019] Zhixiong Nan, Yang Liu, Nanning Zheng, and Song-Chun Zhu. Recognizing unseen attribute-object pair with generative model. In Proceedings of the AAAI Conference on Artificial Intelligence,

    2019

  29. [2019]

    Hsva: Hierarchical semantic-visual adaptation for zero-shot learning

    [Chen et al., 2021] Shiming Chen, Guosen Xie, Yang Liu, Qinmu Peng, Baigui Sun, Hao Li, Xinge You, and Ling Shao. Hsva: Hierarchical semantic-visual adaptation for zero-shot learning. Advances in Neural Information Pro- cessing Systems, 34:16622–16634,

    2021

  30. [2020]

    Prompting language-informed distribution for compositional zero-shot learning

    [Bao et al., 2025] Wentao Bao, Lichang Chen, Heng Huang, and Yu Kong. Prompting language-informed distribution for compositional zero-shot learning. InEuropean Confer- ence on Computer Vision, pages 107–123. Springer,

    2025

  31. [2021]

    An online algorithm for hierarchical phoneme classification

    [Dekel et al., 2004] Ofer Dekel, Joseph Keshet, and Yoram Singer. An online algorithm for hierarchical phoneme classification. In International workshop on machine learning for multimodal interaction , pages 146–158. Springer,

    2004

  32. [2022]

    Learning attention as disentangler for composi- tional zero-shot learning

    [Hao et al., 2023] Shaozhe Hao, Kai Han, and Kwan-Yee K Wong. Learning attention as disentangler for composi- tional zero-shot learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 15315–15324,

    2023

  33. [2023]

    Troika: Multi-path cross-modal traction for compositional zero- shot learning

    [Huang et al., 2024] Siteng Huang, Biao Gong, Yutong Feng, Min Zhang, Yiliang Lv, and Donglin Wang. Troika: Multi-path cross-modal traction for compositional zero- shot learning. In Proceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition , pages 240...

    2024

  34. [2024]

    Procc: Progressive cross-primitive compatibility for open-world compositional zero-shot learning

    [Huo et al., 2024] Fushuo Huo, Wenchao Xu, Song Guo, Jingcai Guo, Haozhao Wang, Ziming Liu, and Xiaocheng Lu. Procc: Progressive cross-primitive compatibility for open-world compositional zero-shot learning. In Proceed- ings of the AAAI Conference on Artificial Intelligence ,

    2024

  35. [2025]

    Analysis of representations for domain adaptation

    [Ben-David et al., 2006] Shai Ben-David, John Blitzer, Koby Crammer, and Fernando Pereira. Analysis of representations for domain adaptation. Advances in neural information processing systems, 19,

    2006

Pith tools

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