MTCAE-DFER: Multi-Task Cascaded Autoencoder for Dynamic Facial Expression Recognition

Kaida Wu; Ou Bai; Peihao Xiang

REVIEW 4 major objections 6 minor 31 references

MTCAE-DFER: Multi-Task Cascaded Autoencoder for Dynamic Facial Expression Recognition

T0 review · 4 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 The paper argues that cascading dynamic face detection, landmark localization, and expression recognition through a shared VideoMAE encoder with task-level ViT decoders beats single-task and non-cascaded multi-task baselines on RAVDESS…

desk verdict Clean internal ablation, but the SOTA claim rests on unverified quoted numbers—worth a serious referee if the protocol gap is fixed. read the letter →

arxiv 2412.18988 v2 pith:QPK6OFET submitted 2024-12-25 cs.CV cs.LGcs.MM

Peihao Xiang , Kaida Wu , Ou Bai This is my paper · ORCID

classification cs.CVcs.LGcs.MM

keywords dynamicfacialexpressionrecognitionmulti-tasklearningcascadedautoencoderVisionTransformerdecoderVideoMAEglobal-localfeatureinteractionfacedetectionlandmark

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 dynamic facial expression recognition is best treated as the last stage of a cascade of related face tasks. Its MTCAE-DFER model shares one VideoMAE encoder across dynamic face detection, dynamic face landmark, and expression recognition, and connects the tasks through a plug-and-play Cascaded ViT Decoder in which the shared encoder supplies K and V while the previous task's decoder output supplies Q. On RAVDESS, CREMA-D, and MEAD, the paper reports weighted average recall of 83.69, 85.03, and 88.44, beating the prior multi-task self-supervised baseline MTFormer by about 4 points on each dataset and the single-task baseline by 7 to 9 points. If the comparison holds, the value is a way to inject fine-grained facial geometry into video emotion recognition without a separate offline pipeline.

What carries the argument

The Cascaded ViT Decoder is the load-bearing module. It is a compact Transformer decoder, configured here with 5 layers and 3 attention heads at feature dimension 512, whose multi-head attention is re-pointed so that K and V come from the normalized dense output of the shared VideoMAE encoder and Q comes from the previous task's decoder output (or from the encoder itself for the first task). Q updates through recursion while K and V stay fixed, so each stage mixes global video features with local task features. The cascade is written as Y1 = ViTDecoder(F,F), Y2 = ViTDecoder(Y1,F), and Y3 = ViTDecoder(Y2,F), where F is the encoder output; this chaining is what turns parallel multi-task decoders into a task cascade.

What would settle it

Retrain MAE-DFER and MTFormer under the exact protocol used for MTCAE-DFER — 16 frames per clip via average downsampling, MTCNN-generated auxiliary labels, 5-fold subject-independent cross-validation, and matched training budgets — and compare WAR; if the gaps in Tables 2-4 shrink to noise, the central claim fails. A second check: replace the cascaded Q input of the DFER decoder with a fixed non-cascaded local feature of the same dimension; if WAR does not drop, the cascade order is not the cause.

Watch

Extended reading notes

Core claim

The central claim is that cascading three face-related tasks through a shared encoder and task-level ViT decoders consistently outperforms both single-task learning and non-cascaded multi-task learning across all three datasets. The mechanism is global-local feature interaction: at each stage, the global dynamic representation from the VideoMAE encoder is entered as K and V, and the local output of the previous stage is entered as Q, so the expression head receives features already filtered by face detection and landmark localization. In the paper's reported numbers, the cascaded model reaches WAR 83.69 on RAVDESS (7-class), 85.03 on CREMA-D (6-class), and 88.44 on MEAD (8-class), improving on the previous best multi-task self-supervised model MTFormer by 4.01, 4.29, and 3.70 points respectively.

Load-bearing premise

The load-bearing premise is that the comparison numbers quoted from other papers were produced under protocols compatible enough with the new model's subject splits, frame sampling, and training setup; if those protocols differ, the reported gains could be artifacts of experimental setup rather than the cascaded architecture.

Editorial extensions

If this is right

Expression recognition can be improved by consuming features already shaped by face detection and landmark localization, not only raw global video tokens.
The Cascaded ViT Decoder is plug-and-play: attaching it to a shared VideoMAE encoder and cascading the three tasks adds 1.52 to 2.26 WAR points over the non-cascaded ViT decoder MTL variant on the three datasets.
Even with MLP decoders, multi-task learning with the shared encoder beats single-task fine-tuning by at least 1.32 to 2.40 WAR points, so the MTL formulation itself contributes to generalization.
The same task-chain recipe can be transferred to other related-task video problems, and the authors propose combining it with emotion-based text generation and multi-modal encoders.

Reading between the lines

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

The cascade's success may depend on the proxy labels: because MTCNN generates the detection and landmark supervision, a clean test would reorder or randomly relabel the auxiliary tasks to see whether the gain comes from task order or simply from added supervision.
The reported margin over prior models is measured against numbers quoted from other papers; re-running those baselines under identical 16-frame downsampling, subject-independent 5-fold splits, and training budgets could yield smaller margins than the tables show.
The same global-K/V plus local-Q design could apply to other video-understanding chains, such as person detection to pose estimation to action recognition, where the shared encoder provides global tokens and the earlier task provides local keypoints.

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.

Desk Editor's Note

read the letter

The internal ablation is clean and the cascade gives a consistent small gain, but the SOTA claim does not hold up as written because Tables 2-4 are quoted numbers from other papers, not protocol-matched reruns. That's the one thing you should remember.

What's actually new: MTCAE-DFER chains three ViT decoders over a shared VideoMAE encoder—face detection output becomes the Q for the landmark decoder, whose output becomes the Q for the expression decoder, while K and V stay fixed from the encoder. That cascaded global-local attention is a specific arrangement I haven't seen in prior DFER work, and the paper describes it clearly. Credit: the ablation (Table 1) compares six structures and the ordering is consistent across RAVDESS, CREMA-D, MEAD. The cascade beats the best non-cascaded ViT decoder by 1.5-2.3 WAR points, and the ViT decoder consistently beats MLP decoders by 3.8-4.9 points. Those are believable, internally consistent results.

Soft spots, in proportion. First, the headline 'consistently surpasses current SOTA models across all datasets' is only supported by Tables 2-4, where MAE-DFER, MTFormer, SVFAP etc. are quoted from their own papers. Nothing says those models were run under the same 5-fold subject-independent splits, 16-frame average downsampling, MTCNN-generated labels, VideoMAE-L initialization, and AdamW settings. Differences of 3.7-4.3 WAR over MTFormer are within reach of protocol changes. That claim needs reruns or a star in the table saying 'as reported in original paper,' with a protocol table to show comparability. Second, Section 4.3 describes the STL head as 'only Fully Connected layer,' while Section 4.2 defines task-specific heads as Norm+Pool+FC. If the STL baseline really was a plain FC, the 8.8-9.25 WAR gain attributed to the cascade is partly a head upgrade. That ambiguity should be fixed—ideally by rerunning STL with the same head. Minor: no error bars, no code, and no cascaded-MLP ablation, though the cascade effect is already isolated against non-cascaded ViT decoders; the missing row would be informative, not load-bearing.

Overall, the paper is honest and methodologically clear, with limitations acknowledged. It's an incremental architecture change, not a breakthrough. I'd trust the internal comparison for direction, but I wouldn't put it on a reference list until the SOTA claim is verified. If the authors supply protocol-matched baselines and clarify the STL head, it's a solid contribution to affective computing. I'd accept this for peer review, with a request for those experiments. Reading group? Maybe, as a case study of why internal ablations can't carry an external SOTA claim.

Referee Report

4 major / 6 minor

Summary. The manuscript proposes MTCAE-DFER, a multi-task cascaded autoencoder framework for dynamic facial expression recognition. It uses a shared VideoMAE encoder and three cascaded ViT decoders for face detection, face landmark, and expression recognition, where the decoder output of each previous task serves as the query for the next task's cross-attention with the shared encoder's global features. The paper reports ablation studies over six architecture variants on RAVDESS, CREMA-D, and MEAD, and compares against SOTA methods, claiming consistent improvements. The central claims are that cascading related facial tasks improves DFER accuracy and that the proposed cascaded ViT decoder enables global-local feature interaction.

Significance. If the reported results hold under controlled comparison, the paper makes a useful contribution: it introduces a plug-and-play cascaded ViT decoder for multi-task learning in DFER and provides a self-consistent internal ablation showing that cascaded ViT decoding outperforms non-cascaded variants. The method is clearly described and the ablation ordering is coherent. The main limitations are the uncontrolled baseline comparisons and the absence of statistical detail, which currently leave the core SOTA claim unverified.

major comments (4)

[§4.4.2, Tables 2–4] The state-of-the-art comparison is the central evidence for the claim in §4.5 that MTCAE-DFER 'consistently surpasses the current SOTA models across all datasets.' However, Tables 2–4 quote WAR/UAR values from prior publications without any statement that those baselines were evaluated under the same protocol as the proposed model. The paper's own protocol includes 5-fold subject-independent cross-validation, 16-frame average downsampling to 224×224, MTCNN-generated face/landmark labels, VideoMAE-L pretrained initialization, AdamW with lr=1e-3, and specific loss weights (0.5/0.5/1.5). Published results for MAE-DFER, MTFormer, SVFAP, and MTL-ER may use different frame sampling, input resolutions, subject splits, or training budgets, and protocol differences of this kind can easily account for the 3.7–4.3 percentage-point WAR gaps over MTFormer reported here. The authors should either re-run the baselines in their own pipeline or provide explicit, verifiable documentation that the quoted numbers were produced under identical protocols; without this, the SOTA claim is not established.
[§4.3 vs §4.2] The STL ablation baseline is described in §4.3 as 'the VideoMAE encoder and the task-specific head with only Fully Connected layer,' whereas the task-specific heads used in the MTL variants are defined in §4.2 as including Normalization, Pooling, and Fully Connected layers. If the STL head lacks pooling and normalization, it is a weaker predictor, and the large STL-to-MTL gains in Table 1 (for example, 9.25 WAR points on RAVDESS) may partly reflect a head-architecture mismatch rather than the effect of MTL or cascading. The STL baseline should use the same Norm+Pool+FC head as the MTL variants to make the comparison controlled.
[Table 1, §4.5] The ablation study only compares the full three-task cascade against STL and non-cascaded MTL variants; it does not include configurations with only face detection or only face landmark as auxiliary tasks, nor does it test alternative task orderings. Consequently, the paper's stated goal of exploring the 'impact of dynamic face detection and dynamic face landmark on dynamic facial expression recognition' is not directly supported: one cannot tell from Table 1 whether both auxiliary tasks contribute, whether one dominates, or whether the specific order detection→landmark→expression is important. Adding these ablations would substantially strengthen the causal claims in §4.5.
[Tables 1–4] All experiments appear to be single runs, with no error bars, per-fold results, or significance tests. Several of the key differences are small—for example, the cascade over non-fully-shared ViT decoder improves WAR by 1.52 points on RAVDESS, 1.36 on CREMA-D, and 1.87 on MEAD in Table 1. Without an estimate of run-to-run variance (at least 3–5 seeds with mean and standard deviation, or paired per-fold comparisons), these differences, and the corresponding SOTA margins, cannot be distinguished from noise. The authors should report variance or per-fold breakdowns.

minor comments (6)

[Abstract and §4.5] The abstract states that the model 'alleviates overfitting of complex large model,' but the paper provides no overfitting analysis (e.g., train/validation gap or regularization comparison). Either provide evidence or remove the claim.
[§4.2, Algorithm 1] The dimension of the Dense layer that produces K and V is not specified in the algorithm; §3.3 later gives the ViT Decoder feature dimension as 512. Please state this dimension in the algorithm description.
[§4.4.2] The text reports differences in percentage points (e.g., '20.70%') but should say 'percentage points' to avoid confusion with relative improvement.
[References] Reference [30] is a millimeter-wave radar gesture recognition paper and does not support the sentence in the Introduction citing 'ResNet18 + LSTM' for dynamic facial expression recognition. Please replace it with a relevant reference.
[Throughout] The dataset name is rendered as 'RA VDESS' and the metric as 'W AR' with an extra space in several places (for example, §1, §4.1, Table 2). Please fix the spacing.
[§4.2] The hyperparameters (loss weights 0.5/0.5/1.5, decoder depth 5, heads 3) are chosen without a sensitivity study; a small analysis of the loss-weight ratio would help, though this is not essential.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: MTCAE-DFER is an empirical architecture paper whose claims rest on ablations and external comparisons, not on definitions, fitted predictions, or self-citations.

full rationale

I walked the paper's derivation chain and found no step in which a predicted quantity reduces to an input by construction. The central claim in Section 4.5 ('the results of multi-task cascaded learning MTCAE-DFER consistently surpass the current SOTA models across all datasets') is supported by Table 1 ablations and Tables 2-4 comparisons, not by a tautological fit. The Cascaded ViT Decoder (Eq. 4-6) is an architectural design where Q comes from the previous task and K,V come from the shared encoder, but the paper does not define the target metric in terms of this design; it measures WAR/UAR empirically. No parameter is fitted to a subset and then reported as a prediction of a closely related quantity: the auxiliary face-detection and landmark losses (Eq. 7-8) are trained tasks, not disguised evaluation metrics. The reference list contains no works by the present authors, so no self-citation chain is load-bearing. The two concerns noted in the review are not circularity: the SOTA tables quote baselines under possibly different protocols, which is a comparability/correctness risk, and the STL ablation head is described inconsistently ('only Fully Connected layer' in Section 4.3 versus Norm+Pool+FC in Section 3.3), which is an internal-consistency matter that could affect the magnitude of reported gains but does not make any result equivalent to its own input. Therefore the paper is self-contained as an empirical study, with no significant circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claims rely on the transferability of the VideoMAE encoder, the validity of MTCNN-generated auxiliary labels, the unablated task order, and the comparability of external baseline numbers.

free parameters (2)

loss weights = 0.5 (detection), 0.5 (landmark), 1.5 (expression)
Chosen by hand in Section 4.2 without sensitivity analysis; affects the balance of the three tasks in the total loss.
ViT Decoder hyperparameters = 5 layers, 3 heads, feature dimension 512
Stated in Section 4.2 without ablation or justification.

assumptions (4)

domain assumption VideoMAE pre-trained encoder generalizes to RAVDESS, CREMA-D, MEAD
The model inherits the VideoMAE encoder and fine-tunes on these datasets (Section 3.1).
domain assumption MTCNN-generated bounding boxes and landmarks are valid ground truth for auxiliary tasks
Labels are auto-generated per frame (Section 4.2); errors in these labels propagate to the cascade.
ad hoc to paper The task order detection -> landmark -> expression is the correct inductive bias
Hypothesized in Section 3.2; the order is not ablated.
domain assumption Baseline numbers from prior papers were obtained under the same evaluation protocol
Tables 2-4 quote prior results; no protocol reconciliation is reported.

how reviews work

0 comments

Cite this review

Pith. "Pith review of MTCAE-DFER: Multi-Task Cascaded Autoencoder for Dynamic Facial Expression Recognition." pith.science (2026). https://pith.science/paper/QPK6OFET

@misc{pith2026241218988,
  author       = {Pith},
  title        = {Pith review of: MTCAE-DFER: Multi-Task Cascaded Autoencoder for Dynamic Facial Expression Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QPK6OFET}},
  note         = {Machine review of arXiv:2412.18988}
}

read the original abstract

This paper expands the cascaded network branch of the autoencoder-based multi-task learning (MTL) framework for dynamic facial expression recognition, namely Multi-Task Cascaded Autoencoder for Dynamic Facial Expression Recognition (MTCAE-DFER). MTCAE-DFER builds a plug-and-play cascaded decoder module, which is based on the Vision Transformer (ViT) architecture and employs the decoder concept of Transformer to reconstruct the multi-head attention module. The decoder output from the previous task serves as the query (Q), representing local dynamic features, while the Video Masked Autoencoder (VideoMAE) shared encoder output acts as both the key (K) and value (V), representing global dynamic features. This setup facilitates interaction between global and local dynamic features across related tasks. Additionally, this proposal aims to alleviate overfitting of complex large model. We utilize autoencoder-based multi-task cascaded learning approach to explore the impact of dynamic face detection and dynamic face landmark on dynamic facial expression recognition, which enhances the model's generalization ability. After we conduct extensive ablation experiments and comparison with state-of-the-art (SOTA) methods on various public datasets for dynamic facial expression recognition, the robustness of the MTCAE-DFER model and the effectiveness of global-local dynamic feature interaction among related tasks have been proven.

Figures

Figures reproduced from arXiv: 2412.18988 by the authors.

**Figure 1.** Illustration of the differences between the following four frameworks: (a) Autoencoder-Based Single-Task Learning Framework, [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

**Figure 2.** MTCAE-DFER Model Structure 2.2. Multi-Task Learning MTL is mainly used for joint learning of multiple tasks to enhance the generalization ability of the model. Various MTL variants have been developed by designing the shared module to effectively distribute and share information across different tasks, as seen in structures like MTFormer [27]. In this study, we focus on exploring and extending the various MTL fra… view at source ↗

**Figure 3.** Multi-Task Label Data Preprocessing Flow-Chart [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 24 canonical work pages

[1]

Bertasius, H

G. Bertasius, H. Wang, and L. Torresani. Is space-time at- tention all you need for video understanding? In ICML, volume 2, page 4, 2021

work page 2021
[2]

H. Cao, D. G. Cooper, M. K. Keutmann, R. C. Gur, A. Nenkova, and R. Verma. Crema-d: Crowd-sourced emo- tional multimodal actors dataset. IEEE transactions on af- fective computing, 5(4):377–390, 2014

work page 2014
[3]

Carreira and A

J. Carreira and A. Zisserman. Quo vadis, action recognition? a new model and the kinetics dataset. In proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 6299–6308, 2017

work page 2017
[4]

K. Cho. Learning phrase representations using rnn encoder- decoder for statistical machine translation. arXiv preprint arXiv:1406.1078, 2014

arXiv 2014
[5]

J. Dai, K. He, and J. Sun. Instance-aware semantic segmen- tation via multi-task network cascades. In Proceedings of the IEEE conference on computer vision and pattern recog- nition, pages 3150–3158, 2016

work page 2016
[6]

Dosovitskiy

A. Dosovitskiy. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020

arXiv 2010
[7]

J. L. Elman. Finding structure in time. Cognitive science, 14(2):179–211, 1990

work page 1990
[8]

Feichtenhofer, H

C. Feichtenhofer, H. Fan, J. Malik, and K. He. Slow- fast networks for video recognition. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6202–6211, 2019

work page 2019

Show all 31 references

[9]

Ghaleb, M

E. Ghaleb, M. Popa, and S. Asteriadis. Multimodal and tem- poral perception of audio-visual cues for emotion recogni- tion. In 2019 8th international conference on affective com- puting and intelligent interaction (ACII) , pages 552–558. IEEE, 2019

2019
[10]

K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick. Masked autoencoders are scalable vision learners. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 16000–16009, 2022

2022
[11]

Hochreiter

S. Hochreiter. Long short-term memory. Neural Computa- tion MIT-Press, 1997

1997
[12]

Huang, C

Q. Huang, C. Huang, X. Wang, and F. Jiang. Facial expres- sion recognition with grid-wise attention and visual trans- former. Information Sciences, 580:35–54, 2021

2021
[13]

Le Ngwe, K

J. Le Ngwe, K. M. Lim, C. P. Lee, T. S. Ong, and A. Alqah- tani. Patt-lite: lightweight patch and attention mobilenet for challenging facial expression recognition. IEEE Access, 2024

2024
[14]

S. R. Livingstone and F. A. Russo. The ryerson audio- visual database of emotional speech and song (ravdess): A dynamic, multimodal set of facial and vocal expressions in north american english. PloS one, 13(5):e0196391, 2018

2018
[15]

Loshchilov

I. Loshchilov. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017

2017 arXiv
[16]

F. Ma, B. Sun, and S. Li. Spatio-temporal transformer for dynamic facial expression recognition in the wild. arXiv preprint arXiv:2205.04749, 2022

2022 arXiv
[17]

F. Ma, B. Sun, and S. Li. Logo-former: Local-global spatio- temporal transformer for dynamic facial expression recogni- tion. In ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 1–5, 2023

2023
[18]

Minaee, M

S. Minaee, M. Minaei, and A. Abdolrashidi. Deep-emotion: Facial expression recognition using attentional convolutional network. Sensors, 21(9):3046, 2021

2021
[19]

Ouyang, S

X. Ouyang, S. Xu, C. Zhang, P. Zhou, Y . Yang, G. Liu, and X. Li. A 3d-cnn and lstm based multi-task learning architec- ture for action recognition. IEEE Access, 7:40757–40770, 2019

2019
[20]

L. Su, C. Hu, G. Li, and D. Cao. Msaf: Multimodal split attention fusion. arXiv preprint arXiv:2012.07175, 2020

2012 arXiv
[21]

L. Sun, Z. Lian, B. Liu, and J. Tao. Mae-dfer: Efficient masked autoencoder for self-supervised dynamic facial ex- pression recognition. In Proceedings of the 31st ACM In- ternational Conference on Multimedia , pages 6110–6121, 2023

2023
[22]

L. Sun, Z. Lian, K. Wang, Y . He, M. Xu, H. Sun, B. Liu, and J. Tao. Svfap: Self-supervised video facial affect perceiver. IEEE Transactions on Affective Computing, 2024

2024
[23]

Z. Tong, Y . Song, J. Wang, and L. Wang. Videomae: Masked autoencoders are data-efficient learners for self-supervised video pre-training. Advances in neural information process- ing systems, 35:10078–10093, 2022

2022
[24]

D. Tran, L. Bourdev, R. Fergus, L. Torresani, and M. Paluri. Learning spatiotemporal features with 3d convolutional net- works. In Proceedings of the IEEE international conference on computer vision, pages 4489–4497, 2015

2015
[25]

A. Vaswani. Attention is all you need. Advances in Neural Information Processing Systems, 2017

2017
[26]

K. Wang, Q. Wu, L. Song, Z. Yang, W. Wu, C. Qian, R. He, Y . Qiao, and C. C. Loy. Mead: A large-scale audio-visual dataset for emotional talking-face generation. In European Conference on Computer Vision , pages 700–717. Springer, 2020

2020
[27]

X. Xu, H. Zhao, V . Vineet, S.-N. Lim, and A. Torralba. Mt- former: Multi-task learning via transformer and cross-task reasoning. In European Conference on Computer Vision , pages 304–321. Springer, 2022

2022
[28]

Zhang, Z

K. Zhang, Z. Zhang, Z. Li, and Y . Qiao. Joint face detection and alignment using multitask cascaded convolutional net- works. IEEE signal processing letters , 23(10):1499–1503, 2016

2016
[29]

Zhang, C

T. Zhang, C. Liu, X. Liu, Y . Liu, L. Meng, L. Sun, W. Jiang, F. Zhang, J. Zhao, and Q. Jin. Multi-task learning framework for emotion recognition in-the-wild. In European Confer- ence on Computer Vision, pages 143–156. Springer, 2022

2022
[30]

Zhang, L

Y . Zhang, L. Peng, G. Ma, M. Man, and S. Liu. Dynamic ges- ture recognition model based on millimeter-wave radar with resnet-18 and lstm. Frontiers in Neurorobotics, 16:903197, 2022

2022
[31]

Zhao and Q

Z. Zhao and Q. Liu. Former-dfer: Dynamic facial expression recognition transformer. In Proceedings of the 29th ACM International Conference on Multimedia, pages 1553–1561, 2021

2021

Pith tools

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

[1] [1]

Bertasius, H

G. Bertasius, H. Wang, and L. Torresani. Is space-time at- tention all you need for video understanding? In ICML, volume 2, page 4, 2021

work page 2021

[2] [2]

H. Cao, D. G. Cooper, M. K. Keutmann, R. C. Gur, A. Nenkova, and R. Verma. Crema-d: Crowd-sourced emo- tional multimodal actors dataset. IEEE transactions on af- fective computing, 5(4):377–390, 2014

work page 2014

[3] [3]

Carreira and A

J. Carreira and A. Zisserman. Quo vadis, action recognition? a new model and the kinetics dataset. In proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 6299–6308, 2017

work page 2017

[4] [4]

K. Cho. Learning phrase representations using rnn encoder- decoder for statistical machine translation. arXiv preprint arXiv:1406.1078, 2014

arXiv 2014

[5] [5]

J. Dai, K. He, and J. Sun. Instance-aware semantic segmen- tation via multi-task network cascades. In Proceedings of the IEEE conference on computer vision and pattern recog- nition, pages 3150–3158, 2016

work page 2016

[6] [6]

Dosovitskiy

A. Dosovitskiy. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020

arXiv 2010

[7] [7]

J. L. Elman. Finding structure in time. Cognitive science, 14(2):179–211, 1990

work page 1990

[8] [8]

Feichtenhofer, H

C. Feichtenhofer, H. Fan, J. Malik, and K. He. Slow- fast networks for video recognition. In Proceedings of the IEEE/CVF international conference on computer vision, pages 6202–6211, 2019

work page 2019