Pith. sign in

REVIEW 3 major objections 5 minor 42 references

Dual-view Spatio-Temporal Feature Fusion with CNN-Transformer Hybrid Network for Chinese Isolated Sign Language Recognition

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper's central claim is that a dual-view dataset covering the full Chinese national sign vocabulary, combined with a simple prediction-time fusion of two independently trained view models, substantially improves isolated sign…

desk verdict The NationalCSL-DP dataset is a genuinely valuable resource, but the paper's empirical claims are under-specified because the train/test split protocol is never stated. read the letter →

arxiv 2506.06966 v1 pith:S3XUM7UT submitted 2025-06-08 cs.CV

Siyuan Jing , Guangxue Wang , Haoyang Zhai , Qin Tao , Jun Yang , Bing Wang , Peng Jin This is my paper · ORCID
classification cs.CV
keywords isolatedsignlanguagerecognitiondual-viewdatasetChineseCNN-transformerfeaturefusionNationalCSL-DPhandocclusion
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 presents NationalCSL-DP, a dual-view Chinese isolated sign language recognition dataset containing 134,140 videos of 6,707 glosses recorded from front and left cameras, and claims it is the first dataset to cover the complete Chinese national sign vocabulary while providing aligned two-view videos. It further argues that feeding both views to a recognition model improves accuracy over either single view, and that the best way to exploit the two views is a 'plus fusion' strategy: train a front-view model and a left-view model separately, then sum their predictions at inference. On the largest subset the proposed CNN-transformer baseline reaches 69.61% top-1 accuracy, whereas early and late fusion of the two views fail to beat single-view performance. A sympathetic reader would care because the dataset supplies a vocabulary-complete, occlusion-relevant benchmark for sign language recognition and a strong baseline to compare against.

What carries the argument

The load-bearing mechanism is the dataset-plus-fusion pipeline. NationalCSL-DP was built by deduplicating the 8,124-word Chinese national sign vocabulary to 6,707 distinct glosses (dropping glosses with identical hand motions, and tagging homographs and regional variants with suffixes such as '1-1' and '2-1'), then recording ten fluent signers with two synchronized RGB cameras at 1920×1080 and 50 fps from the front and left sides. The recognition baseline extracts per-frame 512-dimensional features with a ResNet-34, models temporal dependencies with a 4-layer, 8-head transformer encoder, and predicts through an MLP head. 'Plus fusion' trains two such models independently, one per view, and combines their softmax outputs by addition before the final argmax; this is the operation that yields the paper's best results.

What would settle it

Run the plus-fusion CNN-transformer on a signer-disjoint split, training on eight signers and testing on the two unseen signers, and compare dual-view against single-view top-1 accuracy; if the dual-view advantage shrinks or disappears, the claim that the dual-view dataset and plus fusion drive the accuracy gain is not established.

Watch

Extended reading notes

Core claim

The central discovery is that dual-view sign videos are genuinely more informative than single-view videos, but only when the model is allowed to learn each view's temporal dynamics independently and fuse at the decision level. The paper shows this by comparing a skeleton-based model (SL-GCN) and its own CNN-transformer on front-only, left-only, and dual-view inputs across five subsets (200, 500, 1000, 2000, and 6707 glosses). Dual-view plus fusion raises top-1 accuracy on the 6707-gloss subset to 69.61%, up from 64.34% for the front view and 40.99% for the left view with the same backbone; consistent gains appear on every subset. The paper also reports that early fusion (concatenating per-frame CNN features) and late fusion (concatenating transformer outputs) do not improve over single-view accuracy, and on the largest subset they actually hurt, which the authors leave for future work.

Load-bearing premise

The paper assumes its train/test split is fair (for example, signer-independent), but it never describes how the videos were divided, so if the same signer appears in both training and testing the reported dual-view and fusion gains could partly come from signer memorization.

Editorial extensions

If this is right

  • Dual-view input improves top-1 ISLR accuracy for both RGB and skeleton-based models across all five dataset sizes.
  • Plus fusion consistently outperforms early fusion and late fusion, making it a simple and effective baseline for two-view sign recognition.
  • Early and late fusion of dual-view features can underperform single-view input, especially on large vocabularies, indicating that naive feature concatenation does not capture complementary view information.
  • The five-scale dataset arrangement (200 to 6707 glosses) supports studying how recognition difficulty scales with vocabulary size.
  • The CNN-transformer with plus fusion is offered as a strong baseline for future work on NationalCSL-DP.

Reading between the lines

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

  • If the dataset's split is signer-disjoint and the gains persist, NationalCSL-DP becomes a natural testbed for occlusion-robust sign recognition, since the left view often reveals hands hidden in the front view.
  • The failure of early and late fusion hints that the two views' temporal alignments differ; frame-wise feature concatenation may be mismatched. Cross-view attention or view-alignment pretraining is a testable extension the paper does not explore.
  • Because the vocabulary list is the deduplicated national standard and preserves regional variants via suffixes, the dataset could support building sign-language dictionaries and dialect-aware recognition systems, not just isolated word benchmarks.
  • A natural next step is continuous sign language recognition: if sentence-level models also benefit from dual-view plus fusion, the dataset's design could extend beyond isolated glosses.
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

3 major / 5 minor

Summary. The paper introduces NationalCSL-DP, a dual-view Chinese isolated sign language recognition (ISLR) dataset with 134,140 RGB videos from 10 signers covering 6,707 glosses from the Chinese National Sign Language vocabulary, captured simultaneously from front and left views. The dataset is organized into five subsets (200, 500, 1000, 2000, and 6707 glosses). The authors also propose a CNN-transformer baseline and compare early, late, and plus fusion strategies, reporting that dual-view input and plus fusion improve top-1 accuracy, reaching 69.61% on the 6707-gloss subset. The paper claims that this is the first ISLR dataset covering the full national vocabulary and providing aligned dual-view videos.

Significance. If the dataset and the reported results hold up, NationalCSL-DP is a potentially valuable community resource: it is larger in vocabulary than existing Chinese ISLR datasets, it provides two synchronized views to address hand occlusion, and it is released with partition files under a CC-BY license. The proposed plus-fusion baseline is simple and could serve as a useful reference. The paper also has several strengths that should be credited: the dataset construction was supervised by sign-language experts, cross-verification was performed, and the authors provide a concrete download location. However, the empirical claims in the paper are currently not fully supported because the evaluation protocol is underspecified and no uncertainty quantification is given. The missing train/test split description is the most serious issue, since all conclusions about dual-view fusion depend on it.

major comments (3)
  1. [Section 3.2 and Section 5.2] The train/validation/test split protocol is never specified. Section 3.2 states only that K glosses were randomly selected to form the five subsets and that partition files are downloadable; it does not state whether the video-level splits are signer-independent, whether paired front/left videos of the same gloss are kept in the same partition, or what proportion of videos is assigned to training, validation, and test. Section 5.2 mentions early stopping on validation loss but does not define the validation set. With only 10 signers, a random video-level split could put the same signer in both training and test sets, allowing the model to exploit per-signer appearance rather than sign content, which would inflate the dual-view and fusion gains reported in Tables 1 and 2. The authors should state the exact split protocol and, if the splits are not signer-independent, rerun the evaluation under a signer-independent protocol.
  2. [Tables 1 and 2, Section 5.3 and 5.4] All results are reported as single numbers with no error bars, no repeated trials, and no statistical significance test. The text repeatedly says that dual-view input and plus fusion 'significantly increase' accuracy, but on NationalCSL6707 the dual-view over front-view top-1 improvement is 3.26% for SL-GCN and 5.27% for CNN-transformer (Table 1). These differences could be within run-to-run variability for a single seed and a fixed hyperparameter configuration. The authors should report mean and standard deviation over multiple runs, or at least provide a clear statement about stochasticity, and soften the word 'significantly' unless a statistical test is supplied.
  3. [Section 5.3] The claim that the CNN-transformer is a 'strong baseline' is not externally anchored. The only comparison is with SL-GCN on the authors' own new dataset, so there is no previously published result on NationalCSL-DP to calibrate what 'strong' means. This is not by itself a flaw in a dataset paper, but the wording overclaims. The authors should either compare with additional established ISLR methods adapted to this dataset or explicitly limit the claim to 'a baseline for NationalCSL-DP' without the 'strong' qualifier.
minor comments (5)
  1. [Section 3.1] The number of words in the Chinese National Sign Vocabulary is given as 8124 in the Abstract and Introduction but as 8214 in Section 3.1; please reconcile these numbers.
  2. [Equation (2)] Equation (2) defines p_left = S_left(X_front), but it should presumably be S_left(X_left); this typo makes the notation confusing.
  3. [Section 5.1] There are two subsections numbered 5.1 ('Experimental design' and 'Evaluation metric'); the second should be renumbered.
  4. [Equation (8)] Equation (8) is missing the fusion operator between M_front(X_front) and M_left(X_left), and it is unclear whether these outputs are logits or probabilities before the Softmax; please make this explicit.
  5. [Abstract and Conclusions] The video count is inconsistently reported as 134,140 in the Abstract but 131K+ in the Introduction, Contributions, and Conclusions; please standardize.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular dependency found: the paper's central claims are empirical comparisons on a newly introduced dataset, with no self-citations, no fitted parameters renamed as predictions, and no equation reducing to its own input.

full rationale

The paper's derivation chain is empirical rather than formal: it introduces NationalCSL-DP, trains a CNN-transformer baseline with three fusion strategies, and reports top-k accuracy on its own five subsets. There is no equation that reduces to a definition, no fitted constant that is later called a prediction, and no self-citation used as load-bearing support. The 'strong baseline' claim is a relative evaluation against SL-GCN on the same data, not a derived mathematical result, and the fusion comparisons are direct experimental measurements. The unstated train/val/test split protocol is a validity and reproducibility concern, because random video-level splitting with only ten signers could leak signer identity, but that is a correctness risk, not circularity: the reported numbers are not constructed to equal their inputs. The dataset's 'fully covers the CNSV' claim is also qualified by the authors' explicit exclusion of glosses with identical sign motions, which is a preprocessing decision rather than a circular step. Overall, no circular step satisfying the required evidence standard is present.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

No scientific constants are postulated. The paper's central claims rest on data curation assumptions, evaluation protocol assumptions, and standard ML design choices. The most fragile assumption is the unstated split design.

free parameters (4)
  • frame sampling window = 16 frames at spacing 5
    Hand-chosen to give an 80-frame temporal receptive field; affects accuracy.
  • transformer depth = 4 layers
    Hand-chosen; affects temporal modeling capacity.
  • attention heads = 8
    Hand-chosen; affects attention capacity.
  • embedding dimension = 512
    Hand-chosen; affects representation size.
assumptions (5)
  • domain assumption The 10 signers' performances are correct and representative of CNSL.
    Section 3.1 states participants are proficient in CNSL and supervised by experts; no quantitative validation of signing quality is given.
  • domain assumption Deduplicating glosses with identical signs by keeping the first occurrence preserves the national vocabulary coverage.
    Section 3.1: 1,417 glosses removed as duplicates; the retained 6,707 are assumed to cover all distinct signs.
  • domain assumption Front and left camera views provide complementary information that helps resolve occlusions.
    Central motivation in Section 1 and 3; experiments on their own dataset support it but it is not measured against external datasets.
  • ad hoc to paper The train/test partition is unbiased and prevents identity leakage.
    Not stated in the paper; the reported accuracies assume the split is fair. If the same signer appears in both train and test, performance may be inflated.
  • domain assumption ImageNet-pretrained ResNet-34 features transfer to sign video frames.
    Section 5.2 uses a pretrained ResNet-34 without fine-tuning discussion; transfer is assumed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Dual-view Spatio-Temporal Feature Fusion with CNN-Transformer Hybrid Network for Chinese Isolated Sign Language Recognition." pith.science (2026). https://pith.science/paper/S3XUM7UT

@misc{pith2026250606966,
  author       = {Pith},
  title        = {Pith review of: Dual-view Spatio-Temporal Feature Fusion with CNN-Transformer Hybrid Network for Chinese Isolated Sign Language Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S3XUM7UT}},
  note         = {Machine review of arXiv:2506.06966}
}
read the original abstract

Due to the emergence of many sign language datasets, isolated sign language recognition (ISLR) has made significant progress in recent years. In addition, the development of various advanced deep neural networks is another reason for this breakthrough. However, challenges remain in applying the technique in the real world. First, existing sign language datasets do not cover the whole sign vocabulary. Second, most of the sign language datasets provide only single view RGB videos, which makes it difficult to handle hand occlusions when performing ISLR. To fill this gap, this paper presents a dual-view sign language dataset for ISLR named NationalCSL-DP, which fully covers the Chinese national sign language vocabulary. The dataset consists of 134140 sign videos recorded by ten signers with respect to two vertical views, namely, the front side and the left side. Furthermore, a CNN transformer network is also proposed as a strong baseline and an extremely simple but effective fusion strategy for prediction. Extensive experiments were conducted to prove the effectiveness of the datasets as well as the baseline. The results show that the proposed fusion strategy can significantly increase the performance of the ISLR, but it is not easy for the sequence-to-sequence model, regardless of whether the early-fusion or late-fusion strategy is applied, to learn the complementary features from the sign videos of two vertical views.

Figures

Figures reproduced from arXiv: 2506.06966 by the authors.

Figure 1
Figure 1. Figure1 [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure2 [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Proposed CNNtransformer model with different fusion strategies For early fusion 𝐹𝑒𝑎𝑟𝑙𝑦 as shown in [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 42 canonical work pages

  1. [1]

    ACM Transactions on Asian and Low-Resource Language Information Processing, 23(1): 1-31

    Ren T , Yao D, Yang C, Kang X (2024) The influence of Chinese characters on Chinese sign language. ACM Transactions on Asian and Low-Resource Language Information Processing, 23(1): 1-31

    work page 2024

  2. [2]

    In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'23), pp 3210-3219

    Sarhan N, Frintrop S (2023) Unraveling a decade: a comprehensive survey on isolated sign language recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'23), pp 3210-3219

    work page 2023

  3. [3]

    IEEE Transactions on Image Processing, 33: 3486-3495

    Guo L, Xue W, Liu B, Zhang K, Yuan T , Metaxas D (2024) Gloss prior guided visual feature learning for continuous sign language recognition. IEEE Transactions on Image Processing, 33: 3486-3495

    work page 2024

  4. [4]

    In Proceedings of the International Joint Conference on Artificial Intelligence (IJCNN'23), pp 5260-5268

    Yu P , Zhang L, Fu B, Chen Y (2023) Efficient sign language translation with a curriculum-based non- autoregressive decoder. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCNN'23), pp 5260-5268

    work page 2023

  5. [5]

    In Proceedings of the AAA I Conference on Artificial Intelligence (AAAI'24), pp 19643-19651

    Zhao R, Zhang L, Fu B, Hu C, Su J, Chen Y (2024) Conditional variational autoencoder for sign language translation with cross -modal alignment. In Proceedings of the AAA I Conference on Artificial Intelligence (AAAI'24), pp 19643-19651

    work page 2024

  6. [6]

    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'18), pp 7784-7793

    Camgoz N, Hadfield S, Koller O, Ney H, Bowden R (2018) Neural sign language translation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'18), pp 7784-7793

    work page 2018

  7. [7]

    ACM Transactions on Accessible Computing, 8(4): 1-21

    Wang H, Chai X, Hong X, Zhao G, Chen X (2016) Isolated sign language recognition with Grassmann covariance matrices. ACM Transactions on Accessible Computing, 8(4): 1-21

    work page 2016

  8. [8]

    IEEE Transactions on Circuits and Systems for Video Technology, 29(9): 2822-2832

    Huang J, Zhou W, Li H, Li W (2018) Attention -based 3D-CNNs for large-vocabulary sign language recognition. IEEE Transactions on Circuits and Systems for Video Technology, 29(9): 2822-2832

    work page 2018

Show all 42 references
  1. [9]

    IEEE Transactions on Multimedia, 24: 768-779

    Zhou H, Zhou W, Zhou Y , Li H (2021) Spatial -temporal multi -cue network for sign language recognition and translation. IEEE Transactions on Multimedia, 24: 768-779

    2021

  2. [10]

    Neurocomputing, 500: 115-123

    Du Y , Xie P , Wang M, Hu X, Zhao Z, Liu J (2022) Full transformer network with masking future for word-level sign language recognition. Neurocomputing, 500: 115-123

    2022

  3. [11]

    In Proceedings of the International Joint Conference on Artificial Intelligence (IJCNN'23), pp 5968-5976

    Fink J, Poitier P , André M, Meurice L, Frénay B, Cleve A, Meurant L (2023) Sign language -to-text dictionary with lightweight transformer models. In Proceedings of the International Joint Conference on Artificial Intelligence (IJCNN'23), pp 5968-5976

    2023

  4. [12]

    IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(9): 2306-2320

    Koller O, Camgoz N, Ney H, Bowden R (2019) Weakly supervised learning with multi-stream CNN- LSTM-HMMs to discover sequential parallelism in sign language videos. IEEE Transactions on Pattern Analysis and Machine Intelligence, 42(9): 2306-2320

    2019

  5. [13]

    ACM Transactions on Asian and Low -Resource Language Information Processing, 22(1): 1-23

    Rajalakshmi, E, Elakkiya R, Prikhodko A, Grif M, Bakaev M, Saini J, Subram aniyaswamy V (2022) Static and dynamic isolated Indian and Russian sign language recognition with spatial and temporal feature detection using hybrid neural network. ACM Transactions on Asian and Low -R...

    2022

  6. [14]

    In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC/COLING’24), pp 5450-5460

    Hu L, Gao L, Liu Z, Feng W (2024) Dynamic spatial -temporal aggregation for skeleton-aware sign language recognition. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC/COLING’24), pp 5450-5460

    2024

  7. [15]

    In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV'21), pp 3429-3439

    Hosain A, Santhalingam P , Pathak P , Rangwala H, Kosecka J (2021) Hand pose guided 3d pooling for word-level sign language recognition. In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV'21), pp 3429-3439

    2021

  8. [16]

    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 3441-3450

    DeCoster M, VanHerreweghe, M, Dambre J (2021) Isolated sign recognition from RGB video using pose flow and self-attention. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 3441-3450

    2021

  9. [17]

    ACM Transactions on Asian and Low-Resource Language Information Processing, 23(1): 1-19

    Alyami S, Luqman H, Hammoudeh M (2024) Isolated Arabic sign language recognition using a transformer-based model and landmark keypoints. ACM Transactions on Asian and Low-Resource Language Information Processing, 23(1): 1-19

    2024

  10. [18]

    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'23), pp 14890- 14900

    Zuo R, Wei F, Mak B (2023) Natural language-assisted sign language recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'23), pp 14890- 14900

    2023

  11. [19]

    In Procee dings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 3413-3423

    Jiang S, Sun B, Wang L, Bai Y , Li K, Fu Y (2021) Skeleton aware multi -modal sign language recognition. In Procee dings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 3413-3423

    2021

  12. [20]

    In Proceedings of the IE EE/CVF International Conference on Computer Vision (ICCV'21), pp 11087-11096

    Hu H, Zhao W, Zhou W, Wang Y , Li H (2021) SignBERT: pre -training of hand -model-aware representation for sign language recognition. In Proceedings of the IE EE/CVF International Conference on Computer Vision (ICCV'21), pp 11087-11096

    2021

  13. [21]

    IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9): 11221-11239

    Hu H, Zhao W, Zhou W, Li, H (2023) SignBERT+: Hand-model-aware self-supervised pre-training for sign language understanding. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9): 11221-11239

    2023

  14. [22]

    In 2024 IEEE 18th International Conference on Automatic Face and Gesture Recognition (FG'24), pp 1-8

    Kindiroglu A, Kara O, Ozdemir O, Akarun L (2024) Transfer learning for cross -dataset isolated sign language recognition in under-resourced datasets. In 2024 IEEE 18th International Conference on Automatic Face and Gesture Recognition (FG'24), pp 1-8

    2024

  15. [23]

    In Proceedings of the 18th ACM SIGGRAPH International Conference on Virtual-Reality Continuum and its Applications in Industry (VRCAI'22), pp 1-9

    Gao L, Zhu L, Xue S, Wan L, Li P , Feng W (2022) Multi-view fusion for sign language recognition through knowledge transfer learning. In Proceedings of the 18th ACM SIGGRAPH International Conference on Virtual-Reality Continuum and its Applications in Industry (VRCAI'22), pp 1-9

    2022

  16. [24]

    In Proceedings of the 30th British Machine Vision Conference (BMVC’19), pp 100

    Joze H, Koller O (2019) MS-ASL: A large-scale data set and benchmark for understanding American sign language. In Proceedings of the 30th British Machine Vision Conference (BMVC’19), pp 100

    2019

  17. [25]

    In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV'20), pp 1459-1469

    Li D, Rodriguez C, Yu X, Li H (2020) Word -level deep sign language recognition from video: A new large-scale dataset and methods comparison. In Proceedings of the IEEE Winter Conference on Applications of Computer Vision (WACV'20), pp 1459-1469

    2020

  18. [26]

    IEEE Access, 8: 181340-181355

    Sincan O, Keles H (2020) AUTSL: A large scale multi -modal Turkish sign language dataset and baseline methods. IEEE Access, 8: 181340-181355

    2020

  19. [27]

    ACM transactions on multimedia com puting, communications, and applications, 17(3): 1-19

    Hu H, Zhou W, Pu J, Li H (2021) Global -local enhancement network for NMF -aware sign language recognition. ACM transactions on multimedia com puting, communications, and applications, 17(3): 1-19

    2021

  20. [28]

    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 2735-2744

    Duarte A, Palaskar S, Ventura L, Ghadiyaram D, DeHaan K, Metze F, Giro-i-Nieto X (2021) How2Sign: a large-scale multimodal dataset for continuous American sign language. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'21), pp 2735-2744

    2021

  21. [29]

    In Proceedings of the Ann ual Conference on Neural Information Processing Systems (NeurIPS’24)

    Shen X, Du H, Sheng H, Wang S, Chen H, Chen H, Yu X (2024) MM -WLAuslan: multi-view multi- modal word -level Australian sign language recognition dataset. In Proceedings of the Ann ual Conference on Neural Information Processing Systems (NeurIPS’24)

    2024

  22. [30]

    Procedia Computer Science, 192: 971-980

    Nguyen H, Nguyen T (2021) Attention -based network for effective action recognition from multi - view video. Procedia Computer Science, 192: 971-980

    2021

  23. [31]

    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'23), pp 5287-5294

    Ma Y , Yuan L, Abdelraouf A, Han K, Gupta R, Li Z, Wang Z (2023) M2DAR: Multi -view multi-scale driver action recognition with vision transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR'23), pp 5287-5294

    2023

  24. [32]

    In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'21), pp 13359-13368

    Chen Y , Zhang Z, Yuan C, Li B, Deng Y , Hu W (2021) Channel-wise topology refinement graph convolution for skeleton -based action recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'21), pp 13359-13368

    2021

  25. [33]

    In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'21), pp 13413-13422

    Shi L, Zhang Y , Cheng J, Lu H (2021) AdaSGN: Adapting joint number and model size for efficient skeleton-based action recognition. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV'21), pp 13413-13422

    2021

  26. [34]

    In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI'17), pp 4263–4270

    Song S, Lan C, Xing J, Zeng W, Liu J (2017) An end -to-end spa tial-temporal attention model for human action recognition from skeleton data. In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI'17), pp 4263–4270

    2017

  27. [35]

    In Proceedings of the IEEE/CVF international conference on computer vision (ICCV'19), pp 6212 - 6221

    Wang L, Ding Z, Tao Z, Liu Y , Fu Y (2019) Generative multi -view human action recognitio n. In Proceedings of the IEEE/CVF international conference on computer vision (ICCV'19), pp 6212 - 6221

    2019

  28. [36]

    In Proceedings of the European Conference on Computer Vision (ECCV'20), pp 72-90

    Das S, Sharma S, Dai R, Bremond F, Thonnat M (2020) VPN: learning video -pose embedding for activities of daily living. In Proceedings of the European Conference on Computer Vision (ECCV'20), pp 72-90

    2020

  29. [37]

    IEEE Access, 10: 104190- 104201

    Cheng Q, Liu Z, Ren Z, Cheng J, Liu J (2022) Spatial -temporal information aggregation and cross - modality interactive learning for RGB-d-based human action recognition. IEEE Access, 10: 104190- 104201

    2022

  30. [38]

    IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(3): 3522-3538

    Bruce X, Liu Y , Zhang X, Zhong S, Chan K (2022) MMnet: A model-based multimodal network for human action recognition in RGB -d videos. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(3): 3522-3538

    2022

  31. [39]

    In Proceedings of the International Conference on Neural Information Processing Systems (NIPS'18), pp 1262-1272

    Li J, Wong Y , Zhao Q, Kankanhalli M (2018) Unsup ervised learning of view -invariant action representations. In Proceedings of the International Conference on Neural Information Processing Systems (NIPS'18), pp 1262-1272

    2018

  32. [40]

    In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV'23), pp 5573-5583

    Das S, Ryoo M (2023) ViewCLR: Learning self -supervised video representation for unse en viewpoints. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV'23), pp 5573-5583

    2023

  33. [41]

    Pattern Recognition, 113: 107807

    Ji Y , Yang Y , Shen H, Harada T (2021) View-invariant action recognition via unsupervised attention transfer (UANT). Pattern Recognition, 113: 107807

    2021

  34. [42]

    Computer Vision and Image Understanding, 229: 103655

    Bian C, Feng W, Meng F, Wang S (2023) Global-local contrastive multiview representation learning for skeleton-based action recognition. Computer Vision and Image Understanding, 229: 103655

    2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.