Pith. sign in

REVIEW 4 major objections 6 minor 98 references

UniGaze: Towards Universal Gaze Estimation via Large-scale Pre-Training

T0 review · 4 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Masked-autoencoder pre-training on 1.6 million curated in-the-wild face images gives gaze-estimation models lower error on every unseen dataset in the paper's battery, while pre-training designed for semantic tasks fails to help.

desk verdict A solid empirical recipe for MAE pre-training in gaze estimation, but the XGaze test columns are likely inflated: the pre-training set includes novel-view renderings of the same 60 ETH-XGaze identities used for fine-tuning. read the letter →

arxiv 2502.02307 v2 pith:VYNMILSF submitted 2025-02-04 cs.CV

Jiawei Qin , Xucong Zhang , Yusuke Sugano This is my paper · ORCID
classification cs.CV
keywords gazeestimationself-supervisedpre-trainingmaskedautoencodervisiontransformercross-domaingeneralizationdomainfacerepresentationlearningheadposedistribution
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

This paper claims that self-supervised pre-training can substantially improve how well gaze-estimation models transfer to unseen environments — but only when the pre-training is built for gaze's geometric nature rather than for semantic understanding. The authors pre-train a Vision Transformer with Masked Autoencoders (a method that hides random image patches and trains the network to rebuild them) on roughly 1.6 million in-the-wild face images, curated so that faces are normalized on the same geometry downstream gaze models expect, head poses are diverse yet balanced, and identity count is high. Across three evaluation protocols — training on one dataset and testing on others, leaving one dataset out of training, and pooling all datasets into one training set — the resulting UniGaze model reports lower gaze error than CNNs, plain ViTs, semantic self-supervised models, and specialized domain-generalization methods on five public gaze benchmarks. The reason to care: if the recipe holds, gaze systems can be deployed in new environments without collecting new labeled data, and the pre-training principles become a transferable template for other geometric regression tasks.

What carries the argument

The load-bearing mechanism is Masked Autoencoder (MAE) pre-training on a Vision Transformer: 75% of image patches are randomly masked and the encoder–decoder is trained to reconstruct the masked pixels in per-patch normalized space, so the encoder learns facial structure from roughly 1.6 million images without any gaze labels. Around this sits a curated corpus spanning over 260,000 identities — real video and web data (CelebV-Text, VFHQ, VGGFace2) plus synthetic and novel-view-rendered faces (FaceSynthetics, SFHQ-T2I, FFHQ-NV, XGaze-Dense) — all resized to 224x224 through the standard gaze normalization pipeline, with head pose estimated by landmark detection and perspective-n-point to filter extreme angles and measure pose balance. The paper's three discovered prerequisites — normalized input space, balanced wide head-pose coverage, and identity diversity — are what the ablation studies show to be doing the work. A second contribution is procedural: the leave-one-dataset-out and joint-dataset evaluation protocols measure whether one model can serve all domains, which single-dataset cross-evaluation cannot.

What would settle it

Re-run the pipeline with XGaze-Dense removed from pre-training, or replaced by novel-view renders of 60 identities disjoint from ETH-XGaze, and compare the XGaze test error on the strict 20-subject test split used in the supplementary. If the XGaze scores stay about the same, the claim that pre-training learns generalizable face geometry survives; if they degrade substantially, part of the gain comes from identity overlap. A complementary check is to list the test subjects in the Table 4 within-dataset split and verify that none of them belong to the 60 identities rendered into XGaze-Dense.

Watch

Extended reading notes

Core claim

UniGaze's central claim is that masked autoencoding on a large, deliberately curated collection of faces — rather than merely more data or a bigger model — is what unlocks cross-domain gaze estimation. The paper shows that off-the-shelf semantic pre-training (DINO, MoCo-v3, FaRL) and vanilla MAE trained on small gaze data fail to transfer, sometimes doing worse than a plain ResNet-50, and that increasing ViT model size without the right pre-training also fails. Its ablations isolate three load-bearing ingredients: pre-training on normalized face crops that keep the spatial alignment used by downstream gaze models; wide but balanced head-pose distributions so all viewing angles are represented; and over 260,000 distinct identities so facial appearance varies. With these ingredients, UniGaze-H reports the lowest errors in nearly every cross-dataset cell of its tables (for instance 5.57 degrees on ETH-XGaze-to-MPIIFaceGaze versus a best prior of 5.91), and the paper introduces leave-one-dataset-out and joint-dataset protocols as more realistic measures of generalization. The authors' conclusion is that large-scale pre-training helps gaze estimation only when the pre-training task, the input normalization, and the data distribution are aligned with the geometry of gaze.

Load-bearing premise

The reported ETH-XGaze scores assume that pre-training on XGaze-Dense — novel-view renderings of the same 60 subjects used in the joint-dataset fine-tuning and possibly in the within-dataset split — teaches general face geometry instead of memorizing those particular faces; if that assumption fails, the XGaze columns in the main tables overstate the generalization gain.

Editorial extensions

If this is right

  • Practitioners can expect accuracy gains on unseen cameras, identities, and lighting conditions without new labeled gaze data, since UniGaze improves on nearly every cross-dataset transfer direction reported.
  • The benefit scales with pre-training data: error drops consistently as the corpus grows from 25% to 100% of 1.6 million images, with diminishing returns past 75%, making curated data collection a predictable route to better gaze models.
  • Model size pays off only when paired with gaze-appropriate pre-training: an ImageNet-pretrained ViT-Huge often loses to ResNet-50, while UniGaze-Huge is best in most settings, so scaling alone is not a substitute for the right recipe.
  • The two proposed protocols, leave-one-dataset-out and joint-dataset training, give the field a standard way to measure the practical goal of one gaze model serving all domains.
  • Pre-training cannot compensate for a narrow label range at fine-tuning time, since models trained on MPIIFaceGaze still fail on wide-pose datasets, so label diversity in the downstream stage remains essential.

Reading between the lines

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

  • Because each ablation changes pose balance or identity diversity together with total data volume, the paper leaves open which of its three ingredients dominates; a controlled set of synthetic faces with identity count fixed and pose range widened would isolate the mechanism.
  • The same recipe plausibly transfers to neighbouring geometric tasks — head-pose estimation, gaze target detection, or 3D face alignment — where semantic pre-training has similarly underperformed; the paper only demonstrates the effect for gaze direction regression.
  • The pre-training corpus is partly procedural (novel-view and diffusion-generated faces), pointing to a scaling path the paper only hints at: procedurally generating faces with controllable pose and identity could grow the corpus well beyond 1.6 million images, shifting the bottleneck from labeled data to rendering compute.
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

4 major / 6 minor

Summary. The paper proposes UniGaze, a gaze estimation approach that pre-trains a ViT encoder with Masked Autoencoding on a curated corpus of about 1.6 million real and synthetic face images, then fine-tunes on labeled gaze datasets. The authors report consistent cross-dataset improvements over ImageNet pre-trained ViTs, CNN baselines, and prior domain-generalization methods on MPIIFaceGaze, GazeCapture, EYEDIAP, Gaze360, and ETH-XGaze. They also introduce leave-one-dataset-out and joint-dataset evaluation protocols and provide ablations on pre-training data size, composition, normalization, and pose diversity.

Significance. If the reported results hold, the paper provides a valuable practical recipe for gaze-estimation pre-training: use normalized face crops, balanced head-pose distributions, and high identity diversity, rather than generic semantic pre-training. The two proposed evaluation protocols are useful benchmarks for the field. The paper also releases source code and models, which supports reproducibility. The main empirical claim that pre-training improves generalization to unseen gaze domains is, however, weakened by the inclusion of XGaze-Dense in the pre-training corpus, because that dataset is derived from the same identity pool and acquisition pipeline as the ETH-XGaze evaluation set.

major comments (4)
  1. [§3.1, Table 1] The pre-training corpus includes XGaze-Dense, 267,160 novel-view renderings of the 60 ETH-XGaze training subjects, created with the same multi-view reconstruction pipeline used to build ETH-XGaze. The statement in §3.1 that this data is 'equivalent to a generic facial dataset' is not justified: it is unlabeled target-domain data for every XGaze evaluation column in Tables 2, 4, 5, and 6. Since the central claim is generalization to unseen datasets, and the XGaze columns show some of the largest gains (e.g., 16.31 to 11.29 in Table 5), the paper must provide a no-XGaze-Dense ablation to show that the improvement on XGaze is not due to domain or identity memorization.
  2. [§4.4, Table 4] The within-dataset ETH-XGaze evaluation follows the train/test split from 3DGazeNet [73], but the paper does not verify that the test identities in that split are disjoint from the 60 subjects used to generate XGaze-Dense. If they overlap, the within-dataset XGaze number (3.96) is inflated. The authors should either verify disjointness explicitly or re-run the within-dataset evaluation using the standard 60/20 subject split.
  3. [Tables 2-6 and Figure 3] All results are reported as single runs without error bars, confidence intervals, or repeated-seed variance. Given that some margins are small (e.g., Table 4: MPIIFaceGaze 4.07 vs. 4.0, Gaze360 9.44 vs. 9.6), the claim that UniGaze 'consistently' and 'significantly' outperforms baselines is not statistically supported. Adding at least three seeds with standard deviations for the main comparisons would substantially strengthen the paper.
  4. [Table 3] The comparison with AGG, CLIP-Gaze, LG-Gaze, and Gaze-BAR relies entirely on published numbers from works that may use different pre-processing pipelines. The paper re-implements only ResNet-18 and PureGaze, and the supplementary notes minor discrepancies even for those. The claim of 'surpassing SOTA domain generalization methods' is therefore not yet established for the non-re-implemented methods. The authors should either re-implement all compared methods with the same pre-processing or clearly qualify which comparisons are direct and which are indirect.
minor comments (6)
  1. [§3.2] In the sentence 'FaRL-B [96] ... pre-trained on 20 MLAION-Face samples', 'MLAION' appears to be a typo for 'M LAION' (i.e., 20 million LAION-Face samples).
  2. [§4.2, Table 2] The caption defines 'XTest' only in the table footnote; please define it in the main text at first use, and clarify that XGaze Test refers to the 20-subject held-out subset.
  3. [§4.4] The sentence 'For GazeCapture and Gaze360, it is the same as defined in Sec. 4.1' should read 'they are the same as defined in Sec. 4.1' to avoid ambiguity about which split is being referenced.
  4. [Figure 3] The vertical axis label 'Percentage Change from Baseline (%)' is ambiguous; it would be clearer to state that it is the relative error reduction with respect to the 0% (ImageNet pre-trained) model.
  5. [§4.7.1] The statement '0% refers to the ImageNet pre-trained ViT-L' is clear, but the paper should also state whether the 0% baseline undergoes the same fine-tuning protocol as the pre-trained models; otherwise the comparison may not isolate the effect of pre-training.
  6. [Supplementary Table 3] The hyperparameter ranges for color jitter and the probability of grayscale are given only in the supplementary; please also state them briefly in the main text for completeness, since these are part of the proposed pre-training recipe.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the XGaze-Dense overlap is a data-leakage caveat, not a circular derivation.

full rationale

UniGaze's central claim is an empirical recipe: MAE pre-training on a curated 1.6M-image face corpus improves downstream gaze generalization. The derivation chain is not an analytic reduction: pre-training is self-supervised and label-free, downstream heads are trained with a standard L1 gaze loss on external datasets, and comparisons are made against published baselines and re-implementations. No equation defines the reported gains in terms of the training loss or data by construction. The authors' self-citations [61,62] are used to generate novel-view data (FFHQ-NV, XGaze-Dense) and to justify the normalization pipeline [90]; these are method citations, not load-bearing proofs that the pre-training works. The one legitimate concern is not circularity but benchmark hygiene: XGaze-Dense consists of 267,160 novel-view renderings of the 60 ETH-XGaze training identities, so the XGaze test columns in Tables 2, 5, and 6 may benefit from target-domain appearance and pose exposure. This is a leakage or confound question for correctness, not a case where a prediction reduces to its input by construction, and the non-XGaze columns provide independent evidence. No fit parameter is renamed as a prediction, and no uniqueness theorem is imported. Score 1 reflects a clean circularity analysis with a minor data-hygiene caveat.

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

The central claim is empirical and does not rest on a chain of derivation. The key 'free' choices are data curation hyperparameters, and the load-bearing assumptions are about the accuracy of head-pose estimation, the transferability of MAE representations, and the absence of leakage from synthetic pre-training data into evaluation.

free parameters (4)
  • mask ratio = 75%
    MAE masking ratio taken from the original MAE paper, not tuned for gaze; it controls the difficulty of the reconstruction task and the learned representations.
  • head pose filtering threshold = 80 degrees (L2 norm of pitch/yaw)
    Samples with extreme head poses are discarded during pre-training; threshold affects pose diversity and dataset size.
  • sub-sampling rates = every 15 frames (VFHQ), 45 frames (CelebV-Text), 20 images/identity (VGGFace2)
    Hand-chosen to balance redundancy and diversity in pre-training data; affects dataset composition and identity coverage.
  • pre-training epochs = 300 epochs
    Training length for MAE on 1.6M images; more epochs could improve representations but is compute-limited.
assumptions (5)
  • domain assumption Facial landmark detection and PnP give sufficiently accurate head pose for normalizing and filtering pre-training images.
    Section 3.1 uses [7] and [21]; errors in pose estimation propagate to face alignment and pose distribution balancing.
  • domain assumption Self-supervised masked autoencoding on normalized face images learns representations transferable to gaze regression.
    This is the core empirical hypothesis; it is tested by experiments but not proven analytically.
  • domain assumption The gaze labels in the five evaluation datasets are accurate enough to serve as ground truth.
    All error numbers depend on the quality of the public datasets' labels.
  • domain assumption Data normalization [90] preserves the spatial configuration required by downstream gaze models.
    The pre-training input space is aligned to the downstream gaze models; this is inherited from prior work.
  • ad hoc to paper XGaze-Dense and FFHQ-NV are free of identity overlap with the evaluation sets used in each protocol.
    The paper does not verify that the identities or capture conditions in the synthetic pre-training data are disjoint from the downstream training and test sets, particularly for ETH-XGaze.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UniGaze: Towards Universal Gaze Estimation via Large-scale Pre-Training." pith.science (2026). https://pith.science/paper/VYNMILSF

@misc{pith2026250202307,
  author       = {Pith},
  title        = {Pith review of: UniGaze: Towards Universal Gaze Estimation via Large-scale Pre-Training},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VYNMILSF}},
  note         = {Machine review of arXiv:2502.02307}
}
read the original abstract

Despite decades of research on data collection and model architectures, current gaze estimation models encounter significant challenges in generalizing across diverse data domains. Recent advances in self-supervised pre-training have shown remarkable performances in generalization across various vision tasks. However, their effectiveness in gaze estimation remains unexplored. We propose UniGaze, for the first time, leveraging large-scale in-the-wild facial datasets for gaze estimation through self-supervised pre-training. Through systematic investigation, we clarify critical factors that are essential for effective pretraining in gaze estimation. Our experiments reveal that self-supervised approaches designed for semantic tasks fail when applied to gaze estimation, while our carefully designed pre-training pipeline consistently improves cross-domain performance. Through comprehensive experiments of challenging cross-dataset evaluation and novel protocols including leave-one-dataset-out and joint-dataset settings, we demonstrate that UniGaze significantly improves generalization across multiple data domains while minimizing reliance on costly labeled data. source code and model are available at https://github.com/ut-vision/UniGaze.

Figures

Figures reproduced from arXiv: 2502.02307 by the authors.

Figure 1
Figure 1. Leveraging self-supervised pre-training on large-scale facial data, the proposed UniGaze demonstrates strong generalization [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Example of the normalized facial images from different datasets in the pre-training stage. We also draw their head pose [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Effect of MAE pre-training data size on gaze estimation performance. The horizontal axis is the percentage of the pre-training [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (5 more)
Figure 1
Figure 1. Figure 1: Examples comparison of the pixel normalization during the MAE pre-training. The left, middle, and right columns show the [PITH_FULL_IMAGE:figures/full_fig_p014_1.png]
Figure 2
Figure 2. Figure 2: Effect of MAE pre-training dataset composition on [PITH_FULL_IMAGE:figures/full_fig_p014_2.png]
Figure 3
Figure 3. Figure 3: Qualitative results from various in-the-wild video examples. The normalized input images are displayed alongside the original [PITH_FULL_IMAGE:figures/full_fig_p017_3.png]
Figure 4
Figure 4. Figure 4: Qualitative results of in-the-wild video and synthetic video. The normalized input images are displayed alongside the original [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: Qualitative results of examples from the VideoAttentionTarget dataset [ [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

98 extracted references · 73 canonical work pages

  1. [73]

    3dgazenet: Generalizing 3d gaze estimation with weak- supervision from synthetic views

    Evangelos Ververas, Polydefkis Gkagkos, Jiankang Deng, Michail Christos Doukas, Jia Guo, and Stefanos Zafeiriou. 3dgazenet: Generalizing 3d gaze estimation with weak- supervision from synthetic views. In ECCV, pages 387–404. Springer, 2025. 6, 7, 1, 3, 4

    work page 2025

  2. [1]

    L2cs-net: Fine-grained gaze estimation in unconstrained environments

    Ahmed A Abdelrahman, Thorsten Hempel, Aly Khalifa, Ay- oub Al-Hamadi, and Laslo Dinges. L2cs-net: Fine-grained gaze estimation in unconstrained environments. In 2023 8th International Conference on Frontiers of Signal Processing (ICFSP), pages 98–102. IEEE, 2023. 6, 7, 1

    2023

  3. [2]

    Non-intrusive gaze tracking using artificial neural networks

    Shumeet Baluja and Dean Pomerleau. Non-intrusive gaze tracking using artificial neural networks. Proc. NIPS , 6,

  4. [3]

    From feature to gaze: A generaliz- able replacement of linear layer for gaze estimation

    Yiwei Bao and Feng Lu. From feature to gaze: A generaliz- able replacement of linear layer for gaze estimation. InProc. CVPR, pages 1409–1418, 2024. 2, 3, 6

    2024

  5. [4]

    Unsupervised gaze representation learning from multi-view face images

    Yiwei Bao and Feng Lu. Unsupervised gaze representation learning from multi-view face images. InProc. CVPR, pages 1419–1428, 2024. 3

    2024

  6. [5]

    Gaze-based intention estimation: princi- ples, methodologies, and applications in hri

    Anna Belardinelli. Gaze-based intention estimation: princi- ples, methodologies, and applications in hri. ACM Transac- tions on Human-Robot Interaction, 13(3):1–30, 2024. 1

    2024

  7. [6]

    Synthetic faces high quality - text 2 image (sfhq-t2i) dataset, 2024

    David Beniaguev. Synthetic faces high quality - text 2 image (sfhq-t2i) dataset, 2024. 3, 4

    2024

  8. [7]

    How far are we from solving the 2d & 3d face alignment problem? (and a dataset of 230,000 3d facial landmarks)

    Adrian Bulat and Georgios Tzimiropoulos. How far are we from solving the 2d & 3d face alignment problem? (and a dataset of 230,000 3d facial landmarks). In Proc. ICCV,

Show all 98 references
  1. [8]

    Marlin: Masked autoencoder for facial video repre- sentation learning

    Zhixi Cai, Shreya Ghosh, Kalin Stefanov, Abhinav Dhall, Jianfei Cai, Hamid Rezatofighi, Reza Haffari, and Munawar Hayat. Marlin: Masked autoencoder for facial video repre- sentation learning. In Proc. CVPR, pages 1493–1504, 2023. 2, 3

    2023

  2. [9]

    Vggface2: A dataset for recognising faces across pose and age

    Qiong Cao, Li Shen, Weidi Xie, Omkar M Parkhi, and An- drew Zisserman. Vggface2: A dataset for recognising faces across pose and age. In Proc. FG, pages 67–74. IEEE, 2018. 3

    2018

  3. [10]

    Emerg- ing properties in self-supervised vision transformers

    Mathilde Caron, Hugo Touvron, Ishan Misra, Herv ´e J´egou, Julien Mairal, Piotr Bojanowski, and Armand Joulin. Emerg- ing properties in self-supervised vision transformers. In Proc. ICCV, pages 9650–9660, 2021. 2, 5

    2021

  4. [11]

    An empiri- cal study of training self-supervised vision transformers

    Xinlei Chen, Saining Xie, and Kaiming He. An empiri- cal study of training self-supervised vision transformers. In Proc. ICCV, pages 9640–9649, 2021. 2, 5

    2021

  5. [12]

    Gaze estimation using trans- former

    Yihua Cheng and Feng Lu. Gaze estimation using trans- former. In Proc. ICPR, pages 3341–3347. IEEE, 2022. 3, 5, 7

    2022

  6. [13]

    Puregaze: Purify- ing gaze feature for generalizable gaze estimation

    Yihua Cheng, Yiwei Bao, and Feng Lu. Puregaze: Purify- ing gaze feature for generalizable gaze estimation. In Proc. AAAI, pages 436–443, 2022. 2, 3, 5, 6, 4

    2022

  7. [14]

    Appearance-based gaze estimation with deep learning: A re- view and benchmark

    Yihua Cheng, Haofei Wang, Yiwei Bao, and Feng Lu. Appearance-based gaze estimation with deep learning: A re- view and benchmark. PAMI, 2024. 1, 7

    2024

  8. [15]

    Detecting attended visual targets in video

    Eunji Chong, Yongxin Wang, Nataniel Ruiz, and James M Rehg. Detecting attended visual targets in video. In Proc. CVPR, pages 5396–5406, 2020. 4, 7

    2020

  9. [16]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In Proc. CVPR, pages 248–255. Ieee, 2009. 3, 6

    2009

  10. [17]

    An image is worth 16x16 words: Transformers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition ...

    2021

  11. [18]

    Corrupted image modeling for self-supervised vi- sual pre-training

    Yuxin Fang, Li Dong, Hangbo Bao, Xinggang Wang, and Furu Wei. Corrupted image modeling for self-supervised vi- sual pre-training. Proc. ICLR, 2023. 3

    2023

  12. [19]

    Eva: Exploring the limits of masked visual representa- tion learning at scale

    Yuxin Fang, Wen Wang, Binhui Xie, Quan Sun, Ledell Wu, Xinggang Wang, Tiejun Huang, Xinlong Wang, and Yue Cao. Eva: Exploring the limits of masked visual representa- tion learning at scale. In Proc. CVPR, pages 19358–19369,

  13. [20]

    Rt- gene: Real-time eye gaze estimation in natural environments

    Tobias Fischer, Hyung Jin Chang, and Yiannis Demiris. Rt- gene: Real-time eye gaze estimation in natural environments. In Proc. ECCV, pages 334–352, 2018. 2

    2018

  14. [21]

    Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography.Communications of the ACM, 24(6):381–395, 1981

    Martin A Fischler and Robert C Bolles. Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography.Communications of the ACM, 24(6):381–395, 1981. 4

    1981

  15. [22]

    Eyediap: A database for the development and evaluation of gaze estimation algorithms from rgb and rgb-d cameras

    Kenneth Alberto Funes Mora, Florent Monay, and Jean- Marc Odobez. Eyediap: A database for the development and evaluation of gaze estimation algorithms from rgb and rgb-d cameras. In Proc. ETRA, pages 255–258, 2014. 2, 5

    2014

  16. [23]

    Self-supervised facial rep- resentation learning with facial region awareness

    Zheng Gao and Ioannis Patras. Self-supervised facial rep- resentation learning with facial region awareness. In Proc. CVPR, pages 2081–2092, 2024. 2, 3

    2024

  17. [24]

    Automatic gaze analysis: A survey of deep learning based approaches

    Shreya Ghosh, Abhinav Dhall, Munawar Hayat, Jarrod Knibbe, and Qiang Ji. Automatic gaze analysis: A survey of deep learning based approaches. PAMI, 46(1):61–84, 2023. 1

    2023

  18. [25]

    End-to-end video gaze estimation via capturing head-face-eye spatial-temporal interaction context

    Yiran Guan, Zhuoguang Chen, Wenzheng Zeng, Zhiguo Cao, and Yang Xiao. End-to-end video gaze estimation via capturing head-face-eye spatial-temporal interaction context. IEEE Signal Processing Letters, 30:1687–1691, 2023. 6

    2023

  19. [26]

    Application of eye tracking in medicine: A survey, research issues and chal- lenges

    Katarzyna Harezlak and Pawel Kasprowski. Application of eye tracking in medicine: A survey, research issues and chal- lenges. Computerized Medical Imaging and Graphics , 65: 176–190, 2018. 1

    2018

  20. [27]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proc. CVPR, pages 770–778, 2016. 3, 5

    2016

  21. [28]

    Momentum contrast for unsupervised visual repre- sentation learning

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. Momentum contrast for unsupervised visual repre- sentation learning. In Proc. CVPR, pages 9729–9738, 2020. 3

    2020

  22. [29]

    Masked autoencoders are scalable vision learners

    Kaiming He, Xinlei Chen, Saining Xie, Yanghao Li, Piotr Doll´ar, and Ross Girshick. Masked autoencoders are scalable vision learners. In Proc. CVPR, pages 16000–16009, 2022. 2, 3, 4

    2022

  23. [30]

    Rotation-constrained cross-view feature fusion for multi-view appearance-based gaze estimation

    Yoichiro Hisadome, Tianyi Wu, Jiawei Qin, and Yusuke Sugano. Rotation-constrained cross-view feature fusion for multi-view appearance-based gaze estimation. In Proc. WACV, pages 5985–5994, 2024. 3

    2024

  24. [31]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. Proc. NIPS, 33:6840–6851, 2020. 3 9

    2020

  25. [32]

    Masked autoencoders that listen

    Po-Yao Huang, Hu Xu, Juncheng Li, Alexei Baevski, Michael Auli, Wojciech Galuba, Florian Metze, and Christoph Feichtenhofer. Masked autoencoders that listen. Proc. NIPS, 35:28708–28720, 2022. 3

    2022

  26. [33]

    Learning unsupervised gaze repre- sentation via eye mask driven information bottleneck

    Yangzhou Jiang, Yinxin Lin, Yaoming Wang, Teng Li, Bil- ian Ke, and Bingbing Ni. Learning unsupervised gaze repre- sentation via eye mask driven information bottleneck. arXiv preprint arXiv:2407.00315, 2024. 2, 3

    2024 arXiv

  27. [34]

    Redirtrans: Latent-to-latent translation for gaze and head redirection

    Shiwei Jin, Zhen Wang, Lei Wang, Ning Bi, and Truong Nguyen. Redirtrans: Latent-to-latent translation for gaze and head redirection. In Proc. CVPR, pages 5547–5556, 2023. 2

    2023

  28. [35]

    Contrastive represen- tation learning for gaze estimation

    Swati Jindal and Roberto Manduchi. Contrastive represen- tation learning for gaze estimation. In Annual Conference on Neural Information Processing Systems , pages 37–49. PMLR, 2023. 3

    2023

  29. [36]

    Scaling laws for neural language models

    Jared Kaplan, Sam McCandlish, Tom Henighan, Tom B Brown, Benjamin Chess, Rewon Child, Scott Gray, Alec Radford, Jeffrey Wu, and Dario Amodei. Scaling laws for neural language models. arXiv preprint arXiv:2001.08361,

    2001 arXiv

  30. [37]

    Analyzing and improving the image quality of StyleGAN

    Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and improving the image quality of StyleGAN. In Proc. CVPR, 2020. 3, 4

    2020

  31. [38]

    Using eye-tracking in education: review of empirical research and technology

    Fengfeng Ke, Ruohan Liu, Zlatko Sokolikj, Ibrahim Dahlstrom-Hakki, and Maya Israel. Using eye-tracking in education: review of empirical research and technology. Ed- ucational technology research and development, pages 1–36,

  32. [39]

    Gaze360: Physically uncon- strained gaze estimation in the wild

    Petr Kellnhofer, Adria Recasens, Simon Stent, Wojciech Ma- tusik, and Antonio Torralba. Gaze360: Physically uncon- strained gaze estimation in the wild. In Proc. ICCV, pages 6912–6921, 2019. 2, 5

    2019

  33. [40]

    Sapiens: Foundation for human vision mod- els

    Rawal Khirodkar, Timur Bagautdinov, Julieta Martinez, Su Zhaoen, Austin James, Peter Selednik, Stuart Anderson, and Shunsuke Saito. Sapiens: Foundation for human vision mod- els. In Proc. ECCV, pages 206–228. Springer, 2025. 2, 3, 4

    2025

  34. [41]

    Nvgaze: An anatomically-informed dataset for low-latency, near-eye gaze estimation

    Joohwan Kim, Michael Stengel, Alexander Majercik, Shalini De Mello, David Dunn, Samuli Laine, Morgan McGuire, and David Luebke. Nvgaze: An anatomically-informed dataset for low-latency, near-eye gaze estimation. In Proceedings of the 2019 CHI conference on human factors in com...

    2019

  35. [42]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In Proc. ICLR, 2015. 5, 4

    2015

  36. [43]

    Segment any- thing

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C Berg, Wan-Yen Lo, et al. Segment any- thing. In Proc. ICCV, pages 4015–4026, 2023. 2

    2023

  37. [44]

    Weakly-supervised physically unconstrained gaze estimation

    Rakshit Kothari, Shalini De Mello, Umar Iqbal, Wonmin Byeon, Seonwook Park, and Jan Kautz. Weakly-supervised physically unconstrained gaze estimation. In Proc. CVPR, pages 9980–9989, 2021. 3

    2021

  38. [45]

    Eye tracking for everyone

    Kyle Krafka, Aditya Khosla, Petr Kellnhofer, Harini Kan- nan, Suchendra Bhandarkar, Wojciech Matusik, and Antonio Torralba. Eye tracking for everyone. In Proc. CVPR, pages 2176–2184, 2016. 2, 3, 5

    2016

  39. [46]

    Masked autoen- coders for microscopy are scalable learners of cellular biol- ogy

    Oren Kraus, Kian Kenyon-Dean, Saber Saberian, Maryam Fallah, Peter McLean, Jess Leung, Vasudev Sharma, Ayla Khan, Jia Balakrishnan, Safiye Celik, et al. Masked autoen- coders for microscopy are scalable learners of cellular biol- ogy. In CVPR, pages 11757–11768, 2024. 3

    2024

  40. [47]

    Your diffusion model is se- cretly a zero-shot classifier

    Alexander C Li, Mihir Prabhudesai, Shivam Duggal, Ellis Brown, and Deepak Pathak. Your diffusion model is se- cretly a zero-shot classifier. In Proc. ICCV, pages 2206– 2217, 2023. 3

    2023

  41. [48]

    Dreamteacher: Pretraining image backbones with deep generative models

    Daiqing Li, Huan Ling, Amlan Kar, David Acuna, Se- ung Wook Kim, Karsten Kreis, Antonio Torralba, and Sanja Fidler. Dreamteacher: Pretraining image backbones with deep generative models. InProc. ICCV, pages 16698–16708,

  42. [49]

    Uravatar: Universal relightable gaussian codec avatars

    Junxuan Li, Chen Cao, Gabriel Schwartz, Rawal Khirodkar, Christian Richardt, Tomas Simon, Yaser Sheikh, and Shun- suke Saito. Uravatar: Universal relightable gaussian codec avatars. arXiv preprint arXiv:2410.24223, 2024. 4, 6

    2024 arXiv

  43. [50]

    Gazehta: End-to-end gaze target detection with head- target association

    Zhi-Yi Lin, Jouh Yeong Chew, Jan van Gemert, and Xucong Zhang. Gazehta: End-to-end gaze target detection with head- target association. arXiv preprint arXiv:2404.10718, 2024. 2

    2024 arXiv

  44. [51]

    Jitter does matter: Adapting gaze esti- mation to new domains

    Ruicong Liu, Yiwei Bao, Mingjie Xu, Haofei Wang, Yunfei Liu, and Feng Lu. Jitter does matter: Adapting gaze esti- mation to new domains. arXiv preprint arXiv:2210.02082,

    arXiv

  45. [52]

    Pnp- ga+: Plug-and-play domain adaptation for gaze estimation using model variants

    Ruicong Liu, Yunfei Liu, Haofei Wang, and Feng Lu. Pnp- ga+: Plug-and-play domain adaptation for gaze estimation using model variants. PAMI, 2024. 3

    2024

  46. [53]

    From gaze jitter to domain adaptation: Generalizing gaze estimation by ma- nipulating high-frequency components

    Ruicong Liu, Haofei Wang, and Feng Lu. From gaze jitter to domain adaptation: Generalizing gaze estimation by ma- nipulating high-frequency components. IJCV, pages 1–16,

  47. [54]

    Gen- eralizing gaze estimation with outlier-guided collaborative adaptation

    Yunfei Liu, Ruicong Liu, Haofei Wang, and Feng Lu. Gen- eralizing gaze estimation with outlier-guided collaborative adaptation. In Proc. ICCV, pages 3835–3844, 2021. 6

    2021

  48. [55]

    Agisoft metashape

    Agisoft LLC. Agisoft metashape. https : / / www . agisoft.com/, 2024. 3, 4

    2024

  49. [56]

    Head pose-free appearance-based gaze sensing via eye im- age synthesis

    Feng Lu, Yusuke Sugano, Takahiro Okabe, and Yoichi Sato. Head pose-free appearance-based gaze sensing via eye im- age synthesis. In Proc. ICPR, pages 1008–1011. IEEE, 2012. 1

    2012

  50. [57]

    Learning gaze biases with head motion for head pose-free gaze estimation

    Feng Lu, Takahiro Okabe, Yusuke Sugano, and Yoichi Sato. Learning gaze biases with head motion for head pose-free gaze estimation. Image and Vision Computing , 32(3):169– 179, 2014. 1

    2014

  51. [58]

    Gaze-hand align- ment: Combining eye gaze and mid-air pointing for inter- acting with menus in augmented reality

    Mathias N Lystbæk, Peter Rosenberg, Ken Pfeuffer, Jens Emil Grønbæk, and Hans Gellersen. Gaze-hand align- ment: Combining eye gaze and mid-air pointing for inter- acting with menus in augmented reality. Proceedings of the ACM on Human-Computer Interaction , 6(ETRA):1–18,

  52. [59]

    Eye tracking and eye- based human–computer interaction

    P ¨aivi Majaranta and Andreas Bulling. Eye tracking and eye- based human–computer interaction. In Advances in physio- logical computing, pages 39–65. Springer, 2014. 1

    2014

  53. [60]

    Few-shot adaptive gaze estimation

    Seonwook Park, Shalini De Mello, Pavlo Molchanov, Umar Iqbal, Otmar Hilliges, and Jan Kautz. Few-shot adaptive gaze estimation. In Proc. ICCV, pages 9368–9377, 2019. 2, 3, 5 10

    2019

  54. [61]

    Learning-by-novel-view-synthesis for full-face appearance- based 3d gaze estimation

    Jiawei Qin, Takuru Shimoyama, and Yusuke Sugano. Learning-by-novel-view-synthesis for full-face appearance- based 3d gaze estimation. In Proc. CVPRW, pages 4981– 4991, 2022. 2, 3, 4

    2022

  55. [62]

    Domain-adaptive full-face gaze estimation via novel-view-synthesis and feature disentanglement

    Jiawei Qin, Takuru Shimoyama, Xucong Zhang, and Yusuke Sugano. Domain-adaptive full-face gaze estimation via novel-view-synthesis and feature disentanglement. arXiv preprint arXiv:2305.16140, 2023. 3, 4

    2023 arXiv

  56. [63]

    High-resolution image syn- thesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj¨orn Ommer. High-resolution image syn- thesis with latent diffusion models. In Proc. CVPR, pages 10684–10695, 2022. 3

    2022

  57. [64]

    Gazenerf: 3d-aware gaze redirection with neural radiance fields

    Alessandro Ruzzi, Xiangwei Shi, Xi Wang, Gengyan Li, Shalini De Mello, Hyung Jin Chang, Xucong Zhang, and Otmar Hilliges. Gazenerf: 3d-aware gaze redirection with neural radiance fields. In Proc. CVPR, pages 9676–9685,

  58. [65]

    A review of driver gaze estimation and application in gaze behavior understanding

    Pavan Kumar Sharma and Pranamesh Chakraborty. A review of driver gaze estimation and application in gaze behavior understanding. Engineering Applications of Artificial Intel- ligence, 133:108117, 2024. 1

    2024

  59. [66]

    Agent-guided gaze estimation network by two-eye asymmetry exploration

    Yichen Shi, Feifei Zhang, Wenming Yang, Guijin Wang, and Nan Su. Agent-guided gaze estimation network by two-eye asymmetry exploration. In 2024 IEEE International Confer- ence on Image Processing (ICIP) , pages 2320–2326. IEEE,

    2024

  60. [67]

    The effectiveness of mae pre-pretraining for billion- scale pretraining

    Mannat Singh, Quentin Duval, Kalyan Vasudev Alwala, Haoqi Fan, Vaibhav Aggarwal, Aaron Adcock, Armand Joulin, Piotr Doll´ar, Christoph Feichtenhofer, Ross Girshick, et al. The effectiveness of mae pre-pretraining for billion- scale pretraining. In Proc. ICCV, pages 5484–5494, ...

    2023

  61. [68]

    Super-convergence: Very fast training of neural networks using large learn- ing rates

    Leslie N Smith and Nicholay Topin. Super-convergence: Very fast training of neural networks using large learn- ing rates. In Artificial intelligence and machine learning for multi-domain operations applications , pages 369–386. SPIE, 2019. 5, 4

    2019

  62. [69]

    Omnivec: Learn- ing robust representations with cross modal sharing

    Siddharth Srivastava and Gaurav Sharma. Omnivec: Learn- ing robust representations with cross modal sharing. InProc. WACV, pages 1236–1248, 2024. 2, 3

    2024

  63. [70]

    Face-mllm: A large face perception model

    Haomiao Sun, Mingjie He, Tianheng Lian, Hu Han, and Shiguang Shan. Face-mllm: A large face perception model. arXiv preprint arXiv:2410.20717, 2024. 2, 3

    2024 arXiv

  64. [71]

    Appearance-based eye gaze estimation

    Kar-Han Tan, David J Kriegman, and Narendra Ahuja. Appearance-based eye gaze estimation. In Sixth IEEE Workshop on Applications of Computer Vision, 2002.(WACV 2002). Proceedings., pages 191–195. IEEE, 2002. 1

    2002

  65. [72]

    Videomae: Masked autoencoders are data-efficient learn- ers for self-supervised video pre-training

    Zhan Tong, Yibing Song, Jue Wang, and Limin Wang. Videomae: Masked autoencoders are data-efficient learn- ers for self-supervised video pre-training. Proc. NIPS, 35: 10078–10093, 2022. 3

    2022

  66. [74]

    High-fidelity eye animatable neu- ral radiance fields for human face

    Hengfei Wang, Zhongqun Zhang, Yihua Cheng, and Hyung Jin Chang. High-fidelity eye animatable neu- ral radiance fields for human face. arXiv preprint arXiv:2308.00773, 2023. 2

    2023 arXiv

  67. [75]

    Generaliz- ing eye tracking with bayesian adversarial learning

    Kang Wang, Rui Zhao, Hui Su, and Qiang Ji. Generaliz- ing eye tracking with bayesian adversarial learning. In Proc. CVPR, pages 11907–11916, 2019. 1

    2019

  68. [76]

    Con- trastive regression for domain adaptation on gaze estimation

    Yaoming Wang, Yangzhou Jiang, Jin Li, Bingbing Ni, Wen- rui Dai, Chenglin Li, Hongkai Xiong, and Teng Li. Con- trastive regression for domain adaptation on gaze estimation. In Proc. CVPR, pages 19376–19385, 2022. 3

    2022

  69. [77]

    Toward high qual- ity facial representation learning

    Yue Wang, Jinlong Peng, Jiangning Zhang, Ran Yi, Liang Liu, Yabiao Wang, and Chengjie Wang. Toward high qual- ity facial representation learning. In Proceedings of the 31st ACM International Conference on Multimedia, pages 5048– 5058, 2023. 2, 3

    2023

  70. [78]

    Rendering of eyes for eye-shape registration and gaze estimation

    Erroll Wood, Tadas Baltrusaitis, Xucong Zhang, Yusuke Sugano, Peter Robinson, and Andreas Bulling. Rendering of eyes for eye-shape registration and gaze estimation. In Proc. ICCV, pages 3756–3764, 2015. 3

    2015

  71. [79]

    Fake it till you make it: face analysis in the wild using synthetic data alone

    Erroll Wood, Tadas Baltru ˇsaitis, Charlie Hewitt, Sebastian Dziadzio, Thomas J Cashman, and Jamie Shotton. Fake it till you make it: face analysis in the wild using synthetic data alone. In Proc. ICCV, pages 3681–3691, 2021. 3, 4

    2021

  72. [80]

    Denoising diffusion autoencoders are unified self-supervised learners

    Weilai Xiang, Hongyu Yang, Di Huang, and Yunhong Wang. Denoising diffusion autoencoders are unified self-supervised learners. In Proc. ICCV, pages 15802–15812, 2023. 2, 3

    2023

  73. [81]

    Vfhq: A high-quality dataset and benchmark for video face super-resolution

    Liangbin Xie, Xintao Wang, Honglun Zhang, Chao Dong, and Ying Shan. Vfhq: A high-quality dataset and benchmark for video face super-resolution. In Proc. CVPR, pages 657– 666, 2022. 3, 4

    2022

  74. [82]

    Gaze from origin: Learning for generalized gaze estimation by embedding the gaze frontal- ization process

    Mingjie Xu and Feng Lu. Gaze from origin: Learning for generalized gaze estimation by embedding the gaze frontal- ization process. In Proc. AAAI, pages 6333–6341, 2024. 3

    2024

  75. [83]

    Learning a gener- alized gaze estimator from gaze-consistent feature

    Mingjie Xu, Haofei Wang, and Feng Lu. Learning a gener- alized gaze estimator from gaze-consistent feature. In Proc. AAAI, pages 3027–3035, 2023. 2, 3, 5, 6

    2023

  76. [84]

    Nerf-gaze: A head-eye redirection parametric model for gaze estimation.arXiv preprint arXiv:2212.14710,

    Pengwei Yin, Jiawu Dai, Jingjing Wang, Di Xie, and Shil- iang Pu. Nerf-gaze: A head-eye redirection parametric model for gaze estimation.arXiv preprint arXiv:2212.14710,

    arXiv

  77. [85]

    Lg-gaze: Learning geometry-aware continu- ous prompts for language-guided gaze estimation

    Pengwei Yin, Jingjing Wang, Guanzhong Zeng, Di Xie, and Jiang Zhu. Lg-gaze: Learning geometry-aware continu- ous prompts for language-guided gaze estimation. In Proc. ECCV, 2024. 2, 3, 6

    2024

  78. [86]

    Clip-gaze: Towards general gaze estimation via visual- linguistic model

    Pengwei Yin, Guanzhong Zeng, Jingjing Wang, and Di Xie. Clip-gaze: Towards general gaze estimation via visual- linguistic model. In Proc. AAAI, pages 6729–6737, 2024. 2, 3, 6

    2024

  79. [87]

    Celebv-text: A large-scale fa- cial text-video dataset

    Jianhui Yu, Hao Zhu, Liming Jiang, Chen Change Loy, Wei- dong Cai, and Wayne Wu. Celebv-text: A large-scale fa- cial text-video dataset. In Proc. CVPR, pages 14805–14814,

  80. [88]

    Appearance-based gaze estimation in the wild

    Xucong Zhang, Yusuke Sugano, Mario Fritz, and Andreas Bulling. Appearance-based gaze estimation in the wild. In Proc. CVPR, pages 4511–4520, 2015. 1, 2 11

    2015

  81. [89]

    It’s written all over your face: Full-face appearance- based gaze estimation

    Xucong Zhang, Yusuke Sugano, Mario Fritz, and Andreas Bulling. It’s written all over your face: Full-face appearance- based gaze estimation. In Proc. CVPRW, pages 51–60, 2017. 5

    2017

  82. [90]

    Re- visiting data normalization for appearance-based gaze esti- mation

    Xucong Zhang, Yusuke Sugano, and Andreas Bulling. Re- visiting data normalization for appearance-based gaze esti- mation. In Proc. ETRA, 2018. 2, 4, 8

    2018

  83. [91]

    Mpiigaze: Real-world dataset and deep appearance- based gaze estimation

    Xucong Zhang, Yusuke Sugano, Mario Fritz, and Andreas Bulling. Mpiigaze: Real-world dataset and deep appearance- based gaze estimation. IEEE TPAMI, 41(1):162–175, 2019. 2

    2019

  84. [92]

    Eth-xgaze: A large scale dataset for gaze estimation under extreme head pose and gaze variation

    Xucong Zhang, Seonwook Park, Thabo Beeler, Derek Bradley, Siyu Tang, and Otmar Hilliges. Eth-xgaze: A large scale dataset for gaze estimation under extreme head pose and gaze variation. InProc. ECCV, pages 365–381. Springer,

  85. [93]

    Improving domain generalization on gaze estimation via branch-out auxiliary regularization

    Ruijie Zhao, Pinyan Tang, and Sihui Luo. Improving domain generalization on gaze estimation via branch-out auxiliary regularization. arXiv preprint arXiv:2405.01439, 2024. 2, 3, 5, 6, 4

    2024 arXiv

  86. [94]

    Unleashing text-to-image diffusion models for visual perception

    Wenliang Zhao, Yongming Rao, Zuyan Liu, Benlin Liu, Jie Zhou, and Jiwen Lu. Unleashing text-to-image diffusion models for visual perception. In Proc. ICCV, pages 5729– 5739, 2023. 2, 3

    2023

  87. [95]

    Self-learning transformations for improving gaze and head redirection

    Yufeng Zheng, Seonwook Park, Xucong Zhang, Shalini De Mello, and Otmar Hilliges. Self-learning transformations for improving gaze and head redirection. Proc. NIPS, 33: 13127–13138, 2020. 2

    2020

  88. [96]

    General facial representation learning in a visual-linguistic manner

    Yinglin Zheng, Hao Yang, Ting Zhang, Jianmin Bao, Dong- dong Chen, Yangyu Huang, Lu Yuan, Dong Chen, Ming Zeng, and Fang Wen. General facial representation learning in a visual-linguistic manner. In Proc. CVPR, pages 18697– 18709, 2022. 2, 3, 5, 8

    2022

  89. [97]

    Places: A 10 million image database for scene recognition

    Bolei Zhou, Agata Lapedriza, Aditya Khosla, Aude Oliva, and Antonio Torralba. Places: A 10 million image database for scene recognition. IEEE TPAMI, 2017. 4

    2017

  90. [98]

    Deformable one- shot face stylization via dino semantic guidance

    Yang Zhou, Zichong Chen, and Hui Huang. Deformable one- shot face stylization via dino semantic guidance. In Proc. CVPR, pages 7787–7796, 2024. 2 12 Training Data \ Test X test M test GC test Etest G360 test ResNet-50 same-domain 5.25 5.11 3.49 8.51 11.87 leave-one-dataset-out...

    2024

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.