Semi-Supervised Self-Growing Generative Adversarial Networks for Image Recognition

Haoqian Wang; Jun Xu; Lei Zhang; Qionghai Dai; Wangpeng An; Zhiwei Xu

REVIEW 4 major objections 5 minor 79 references

Semi-Supervised Self-Growing Generative Adversarial Networks for Image Recognition

T0 review · 4 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read A self-growing GAN rivals supervised accuracy using only 4% of labels.

desk verdict A plausible but unverified pseudo-label step sits at the center of an otherwise useful empirical combination; worth serious review, not desk rejection. read the letter →

arxiv 1908.03850 v1 pith:6QVLTOAJ submitted 2019-08-11 cs.CV

Haoqian Wang , Zhiwei Xu , Jun Xu , Wangpeng An , Lei Zhang , Qionghai Dai This is my paper · ORCID

classification cs.CV

keywords semi-supervisedlearninggenerativeadversarialnetworkself-growingconvolutionblocktransformationmaximummeandiscrepancyfaceattributerecognitionlabelinferenceimage

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 proposes SGGAN, a semi-supervised generative adversarial network that tries to make unlabeled images substitute for most human labels in recognition tasks. The central claim is that with only about 4% of the facial attribute labels in CelebA, SGGAN reaches accuracy comparable to fully supervised deep networks trained on all labels, and that it beats several semi-supervised GAN baselines on CIFAR-10 and SVHN. The approach rests on three moves: a baby discriminator assigns latent labels to unlabeled images whose predicted probability exceeds $\alpha = 0.98$; a convolution-block-transformation technique grows the network from a baby to a junior to a senior model without wrecking the learned features; and maximum mean discrepancy replaces the usual $\ell^1$ feature-matching distance to stabilize training. A reader should care because the result suggests that abundant unlabeled data can be converted into labels by trusting a classifier's own high-confidence predictions, if the classifier is steadily deepened.

What carries the argument

Three mechanisms carry the argument. Label inference: the discriminator's softmax probability is treated as a confidence score, and any unlabeled image with predicted probability above $\alpha=0.98$ is assigned that predicted label and moved into the training pool. Convolution-block-transformation (CBT): when the network grows from baby to junior to senior, newly added convolution blocks are initialized with Gaussian noise, an identity shortcut is added, and their output is scaled by the adaptive factor $w(t)=1-e^{-t}$, so the shallow network's learned function is preserved while the deeper block gradually takes over. Maximum mean discrepancy (MMD), computed with an inner-product kernel, replaces the $\ell^1$ distance in the generator's feature-matching objective, which the authors find stabilizes training and avoids mode collapse. The discriminator plays two roles at once: adversary in the min-max game and semi-supervised classifier whose confident outputs create new training labels.

What would settle it

Measure the precision of the discriminator's predictions on held-out labeled images at the 0.98 threshold: if a substantial fraction of these high-confidence predictions are wrong, or if a variant that randomly flips a small percentage of pseudo-labels matches the original accuracy, then the label-inference assumption is not what drives the reported gains.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that a single training pipeline can start from a small GAN, use the trained discriminator to pseudo-label unlabeled images with confidence above $\alpha=0.98$, grow the generator and discriminator deeper via convolution-block-transformation, and repeat. On the CelebA face attribute dataset, SGGAN trained with 7,200 labeled images (about 4% of the training set) reaches an average accuracy around 86%, only slightly below the fully supervised LNet+ANet at 87% and above all other compared methods, including Improved GAN and fine-tuned VGG-16 and ResNet-50 at the same label budget. On CIFAR-10 and SVHN, SGGAN reports lower test error than Improved GAN at every label budget tested, with the largest gap at 4,000 CIFAR-10 labels (15.65% versus 18.63%). The authors also report that using MMD as the feature-matching objective lowers the generator's training loss compared with the $\ell^1$ distance, and that the senior generator produces visibly better samples than the baby one.

Load-bearing premise

The method assumes that unlabeled images the discriminator labels with probability above 0.98 are correct often enough that adding them as pseudo-labeled training data improves the classifier rather than injecting noise; if the error rate among those high-confidence predictions is not very low, the self-training loop can amplify errors.

Editorial extensions

If this is right

With only about 4% labeled facial attributes on CelebA, SGGAN matches leading fully supervised methods and beats other semi-supervised GANs on most attributes.
On CIFAR-10 and SVHN, SGGAN lowers test error relative to Improved GAN at each reported label budget, a direct corollary of its claimed label-inference and stabilization gains.
Adding a large external pool of unlabeled images (CelebA) improves LFW-a accuracy by about 6 percentage points, showing the method converts unlabeled volume into accuracy.
The self-growing route through all three generations beats any single-generation model, indicating that depth growth is itself part of the performance gain.
Using MMD instead of $\ell^1$ for feature matching lowers the generator's training loss, which the paper ties to more stable GAN training.

Reading between the lines

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

The fixed threshold $\alpha=0.98$ is chosen on a validation set per attribute; a calibrated or per-class threshold might extend the method to datasets with skewed classes, a test the paper does not run.
CBT is a general weight-transfer recipe: nothing limits it to GANs, so the same grow-deeper-while-preserving-features idea could apply to any deep classifier trained with scarce labels.
The claim that 4% labeled data suffices is demonstrated on face attributes; a natural extension is to test whether the same recipe transfers to domains with less structured or more ambiguous classes, such as medical images or scene recognition.
If the discriminator's probabilities are miscalibrated, the 0.98 threshold may not mean high precision; checking precision-recall on a hold-out set at that threshold would tell whether pseudo-labels are truly clean.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Colleague, here's my read on arXiv:1908.03850. The central claim—4% labeled CelebA attributes gets within a point or two of fully supervised LNet+ANet—is worth taking seriously. If real, that's practically useful. What's genuinely new is the package: self-training via confidence threshold, Net2Net-style network growth with their CBT shortcut-and-scale trick, and MMD as feature matching objective, all in one growing GAN. No single ingredient is new, but the combination and the face-attribute results are.

The paper does some things well. The ablation on growing routes (Table III) shows a clear benefit to baby→junior→senior over single models. The MMD-vs-l1 and CBT-vs-no-CBT comparisons are consistent with their story. The SVHN and CIFAR numbers are competitive with Improved GAN, and face attribute comparisons cover 40 attributes on two datasets.

Soft spots, in order of importance. First, the pseudo-label step is the heart of the 4% claim, and they never verify it. They set alpha=0.98 but don't report how many images pass, what fraction are correct, or any ablation with label inference disabled. The non-monotonic CIFAR row (15.65 at 4000 labels vs 16.51 at 8000) is a yellow flag that pseudo-label feedback may be unstable. Second, the 32x32 datasets: baby D/G are 32x32, but junior/senior are 128x128 and 512x512. There's no explanation of how self-growing applies to CIFAR/SVHN. Third, Eq. (2), the MMD witness function, is written incorrectly—it's a typo that doesn't affect the later correct MMD^2 in Eq. (3), but it's the kind of thing that makes a referee distrust the rest. Fourth, no code, no error bars, and a claimed proof in the introduction that never appears.

The citation pattern is a bit padded with self-citations to denoising papers, but not a big deal. Overall, this is a plausible empirical paper with a load-bearing assumption left untested. It deserves a serious referee: the method is coherent, the experiments are broad, and the central result, if confirmed, is useful. I'd send it to review, but the referee should demand pseudo-label precision analysis and a clear protocol for the 32x32 case.

Referee Report

4 major / 5 minor

Summary. The paper proposes a semi-supervised generative adversarial network called SGGAN for image recognition. The method combines three components: (i) a self-training label-inference step in which unlabeled images whose predicted confidence exceeds a threshold (set to 0.98) are added to the labeled pool; (ii) a self-growing network architecture in which baby, junior, and senior generator/discriminator pairs are trained successively, with weights transferred through a proposed convolution-block-transformation (CBT) technique; and (iii) a feature-matching objective based on maximum mean discrepancy (MMD) instead of the L1 distance used in Improved GANs. Experiments are reported on CIFAR-10, SVHN, CelebA, and LFW-a. The abstract's headline claim is that with only about 4% labeled facial attributes (7,200 images) on CelebA, SGGAN achieves accuracy comparable to fully supervised deep learning methods trained on all labels.

Significance. If the results hold, the paper would make a useful contribution: it would demonstrate that a self-growing GAN with pseudo-labeling can approach supervised performance on face attribute recognition with a very small labeled fraction, and it provides evidence that MMD feature matching and the CBT growth mechanism improve training stability on the tested datasets. The paper has several concrete strengths: experiments span four datasets; ablations isolate the self-growing route, the MMD versus L1 objective, and the CBT versus no-CBT transfer; and comparisons include both semi-supervised GAN baselines and fully supervised VGG/ResNet baselines. The main reservation is that the central mechanism, threshold-based pseudo-label inference, is never directly validated by measuring the correctness of the selected pseudo-labels or by ablating the label-inference step, so the significance of the headline claim is conditional on an assumption that remains untested.

major comments (4)

[III.A.3, III.E.2, IV.B, Table V] The central claim that SGGAN matches supervised accuracy with 4% labeled CelebA attributes rests on the assumption that unlabeled images with predicted probability above alpha=0.98 are labeled correctly by the discriminator. This assumption is asserted in Section III.A.3 and Section III.E.2 but never verified. The paper does not report the precision of the pseudo-labels selected at this threshold, the number of selected images per class, or the class distribution of the selected set, and the ablation study in Section IV.B does not include a run with the label-inference step disabled while MMD and CBT are kept active. Without such an ablation, the gains in Table VII could in principle come from adding systematically biased high-confidence examples rather than from correct label inference. The non-monotonic CIFAR-10 row in Table V (15.65% error with 4,000 labels versus 16.51% with 8,000 labels) is consistent with pseudo-label noise and further motivates this measurement. Please add (i) a direct evaluation of pseudo-label precision and selection statistics on the validation set, (ii) a "no label inference" ablation, and (iii) a per-dataset justification of the 0.98 threshold rather than transferring the CelebA-tuned value to CIFAR-10 and SVHN.
[III.D, Eq. (5)] The discriminator loss in Eq. (5) is not well specified. The text states that xi, gi, and ui represent outputs before the softmax activation, but the supervised term is written as -sum_i label_i * log(x_i); log of a pre-softmax logit is not the cross-entropy loss. If xi is instead intended to be the softmax output, then the statement "before softmax activation" is wrong. This is a load-bearing technical detail because Eq. (5) defines the training objective of the method. Please rewrite Eq. (5) with a clear distinction between logits and posterior probabilities, or state explicitly that the fake-class logit is fixed to zero and that the labeled term uses softmax probabilities.
[III.C, Eq. (2)] The witness function in Eq. (2) is written incorrectly. The standard RKHS witness function for the MMD is f(.) = E_{x~p_data}[K(x, .)] - E_{z~p_z}[K(G(z), .)], a function of a single argument; Eq. (2) as printed mixes the free variable x and the generator variable G(z) inside the kernel in a way that does not define a valid witness function. Although Eq. (3) and Eq. (4) are recognizably the correct squared-MMD expressions, the error in Eq. (2) makes the method description inconsistent and should be corrected.
[III.A, Tables I and II, Section IV.C] The self-growing schedule is not explained for the 32x32 image datasets. Table I lists the junior discriminator as taking 128x128x3 input and the senior discriminator as taking 512x512x3 input, while CIFAR-10 and SVHN experiments use 32x32 images. The paper does not state whether the junior and senior cells are trained at all on these datasets, whether images are resized or upscaled, or whether only the baby cell is used for the results in Tables V and VI. This is essential for reproducibility and for interpreting the claimed benefit of self-growing on CIFAR-10/SVHN. Please specify the exact growth schedule for each dataset or, if only the baby cell is used, say so explicitly.

minor comments (5)

[IV.A (CelebA description)] The dataset split description is internally inconsistent: it first says 19,962 images are used as the testing set and the others as training/validation, but then says a small subset is randomly selected as the training set and the others as the testing set. Please clarify the exact split used for CelebA.
[Figure 10 caption] The caption of Figure 10 says "The loss function of the SGGAN model trained with the CBT v.s. without CBT" but the vertical axis is labeled "Accuracy". The caption should be corrected.
[Algorithm 1] Step 11 of Algorithm 1 ("Initialize a deeper model by using CBT preservation technique") appears visually outside the epoch/batch loops, which makes the timing of the self-growing step ambiguous. Indicate explicitly whether this step is executed after every epoch, after a fixed schedule, or once after the loop.
[III.A.3 and III.E.2] The text says in Section III.A.3 that the threshold is determined on the validation set of CelebA, while Section III.E.2 says the threshold is determined by grid search on the validation set of benchmark datasets. Please state which validation set was used for each dataset and report the grid search range.
[III.E.2] There is a grammatical error in the phrase "Comparing to the than the shallower network" in Section III.E.2; the sentence should be rewritten.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 4%-labeled accuracy claim is an empirical, held-out test result; the pseudo-label threshold is a validation-set hyperparameter.

full rationale

The paper's derivation chain is entirely empirical: SGGAN is constructed from published components (Improved GAN feature matching [52], MMD [14,19], and a Net2Net-inspired convolution-block transformation [8]) and is evaluated on held-out test sets against external baselines, so the central 4%-labeled claim is not an input to the method. The pseudo-label threshold alpha=0.98 is selected on the CelebA validation set (Section III.A.3) and used only as a training hyperparameter; it is not later reported as a prediction, and no equation equates the reported accuracy with this fit. Algorithm 1's feedback loop—the discriminator labels unlabeled examples and is then trained on them—is the algorithm itself, not a derived result, so it does not constitute self-definitional circularity in the sense of a claimed derivation. Self-citations in the reference list are numerous but none is load-bearing; the MMD and feature-matching losses are credited to external prior work [14,19,52]. Two non-circular weaknesses should be flagged: Section I contains a literal '[?]' placeholder in the semi-supervised learning citation list, and the contribution list asserts without proof that 'We prove it is easier to train a model growing from a shallow network to a deep one'; additionally, the paper does not report pseudo-label precision at alpha=0.98, which is a reproducibility and correctness concern rather than a circularity. No circular step was found.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper's central claim rests on standard self-training assumptions, the empirical benefit of depth, a specific kernel choice for MMD, and a possibly non-standard simplification of the Improved GAN loss. The threshold alpha and the growing route are fitted to validation data. No new physical entities are introduced.

free parameters (3)

alpha (label inference threshold) = 0.98
Set via grid search on CelebA validation set (Section III.A.3 and III.E.2); used to decide which unlabeled samples receive pseudo-labels; directly controls the composition of the training set and therefore the reported accuracy.
self-growing route (baby+junior+senior) = baby+junior+senior
Selected as best in Table III by experiments on the gender attribute of CelebA (Section IV.B); this categorical choice affects all subsequent attribute results.
adaptive scaling function w(t) = 1-e^{-t} with t ambiguously defined
Hand-designed in Section III.B; the paper defines t as total iterations per epoch divided by current iteration, which contradicts the claim that w(t) approaches 1 during training.

assumptions (4)

domain assumption Self-training assumption: high-confidence discriminator predictions on unlabeled data are mostly correct and can be used as labels.
Section III.A.3 and III.E.2: the method assigns pseudo-labels to unlabeled samples with probability above alpha=0.98. The paper asserts false positives matter more than false negatives, but provides no theoretical justification.
domain assumption Deeper networks generalize better and can learn from pseudo-labeled data.
Section III.B: the paper motivates self-growing with the empirical success of deeper architectures (AlexNet to VGG to ResNet), assuming depth improves classification.
domain assumption MMD with inner product kernel is a suitable feature matching objective.
Section III.C: the paper uses MMD with the inner product kernel, which is not characteristic; it reduces to matching feature means. The theoretical advantage over l1 distance claimed is not established.
ad hoc to paper The Improved GAN loss can be written with the fake logit fixed to zero.
Equation (5) in Section III.D appears to assume the (k+1)-th logit is 0, which is not the standard Improved GAN formulation; without that assumption the loss is mis-specified.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Semi-Supervised Self-Growing Generative Adversarial Networks for Image Recognition." pith.science (2026). https://pith.science/paper/6QVLTOAJ

@misc{pith2026190803850,
  author       = {Pith},
  title        = {Pith review of: Semi-Supervised Self-Growing Generative Adversarial Networks for Image Recognition},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6QVLTOAJ}},
  note         = {Machine review of arXiv:1908.03850}
}

read the original abstract

Image recognition is an important topic in computer vision and image processing, and has been mainly addressed by supervised deep learning methods, which need a large set of labeled images to achieve promising performance. However, in most cases, labeled data are expensive or even impossible to obtain, while unlabeled data are readily available from numerous free on-line resources and have been exploited to improve the performance of deep neural networks. To better exploit the power of unlabeled data for image recognition, in this paper, we propose a semi-supervised and generative approach, namely the semi-supervised self-growing generative adversarial network (SGGAN). Label inference is a key step for the success of semi-supervised learning approaches. There are two main problems in label inference: how to measure the confidence of the unlabeled data and how to generalize the classifier. We address these two problems via the generative framework and a novel convolution-block-transformation technique, respectively. To stabilize and speed up the training process of SGGAN, we employ the metric Maximum Mean Discrepancy as the feature matching objective function and achieve larger gain than the standard semi-supervised GANs (SSGANs), narrowing the gap to the supervised methods. Experiments on several benchmark datasets show the effectiveness of the proposed SGGAN on image recognition and facial attribute recognition tasks. By using the training data with only 4% labeled facial attributes, the SGGAN approach can achieve comparable accuracy with leading supervised deep learning methods with all labeled facial attributes.

Figures

Figures reproduced from arXiv: 1908.03850 by the authors.

**Figure 2.** The detailed architecture of Baby Generator. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** The detailed architecture of Baby Discriminator. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Convolution Block Transformation (CBT) transfers weights of shallow network to the deeper one. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Samples from the CIFAR-10 dataset [28]. Image Recognition Datasets. In this section, we compare the proposed SGGAN approach with state-of-the-art semisupervised GAN based methods by using the widely used CIFAR-10 dataset [28] and the Street View House Numbers (SVHN) d…

**Figure 8.** Figure 8: Samples from LFWA dataset [24]. B. Ablation Study In this section, we justify the influence of different components in our proposed SGGAN approach on the performance of recognition errors. The aspects we investigate here include the network self-growing route, the obj…

**Figure 7.** Figure 7: Samples from CelebA dataset [40]. Following the experimental settings as the previous work [40], we employ 6, 263 images of 2, 749 peoples as the training set and the other 6, 880 images of 3, 000 peoples as the testing set. When we train the SGGAN model, the labeled i…

**Figure 9.** Figure 9: The loss function of the SGGAN model trained [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

**Figure 10.** Figure 10: The loss function of the SGGAN model trained [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]

**Figure 11.** Figure 11: The generated samples of the baby, junior, and senior generators of the proposed SGGAN approach. [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

79 extracted references · 59 canonical work pages

[1]

A pid controller approach for stochastic optimization of deep networks

Wangpeng An, Haoqian Wang, Qingyun Sun, Jun Xu, Qionghai Dai, and Lei Zhang. A pid controller approach for stochastic optimization of deep networks. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018

work page 2018
[2]

Towards principled methods for training generative adversarial networks

Martin Arjovsky and L ´eon Bottou. Towards principled methods for training generative adversarial networks. In ICLR, 2017

work page 2017
[3]

Wasserstein generative adversarial networks

Mart ´ın Arjovsky, Soumith Chintala, and L ´eon Bottou. Wasserstein generative adversarial networks. In Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017 , pages 214–223, 2017

work page 2017
[4]

Poof: Part-based one-vs.-one features for ﬁne-grained categorization, face veriﬁcation, and attribute estimation

Thomas Berg and Peter Belhumeur. Poof: Part-based one-vs.-one features for ﬁne-grained categorization, face veriﬁcation, and attribute estimation. In CVPR, 2013

work page 2013
[5]

Describing people: A poselet-based approach to attribute classiﬁcation

Lubomir Bourdev, Subhransu Maji, and Jitendra Malik. Describing people: A poselet-based approach to attribute classiﬁcation. In ICCV, 2011

work page 2011
[6]

Semi- Supervised Learning

Olivier Chapelle, Bernhard Schlkopf, and Alexander Zien. Semi- Supervised Learning. The MIT Press, 1st edition, 2010

work page 2010
[7]

Chen and W

K. Chen and W. Tao. Convolutional regression for visual tracking. IEEE Transactions on Image Processing , PP(99):1–1, 2018

work page 2018
[8]

Goodfellow, and Jonathon Shlens

Tianqi Chen, Ian J. Goodfellow, and Jonathon Shlens. Net2net: Accel- erating learning via knowledge transfer. In ICLR, 2016

work page 2016

Show all 79 references

[9]

Semi-supervised multimodal deep learning for rgb-d object recognition

Yanhua Cheng, Xin Zhao, Rui Cai, Zhiwei Li, Kaiqi Huang, and Yong Rui. Semi-supervised multimodal deep learning for rgb-d object recognition. In IJCAI, 2016

2016
[10]

Semi-supervised learning of facial attributes in video

Neva Cherniavsky, Ivan Laptev, Josef Sivic, and Andrew Zisserman. Semi-supervised learning of facial attributes in video. In ECCV, 2010

2010
[11]

Semi-supervised sequence learning

Andrew M Dai and Quoc V Le. Semi-supervised sequence learning. In C. Cortes, N. D. Lawrence, D. D. Lee, M. Sugiyama, and R. Garnett, editors, Advances in Neural Information Processing Systems 28 , pages 3079–3087. Curran Associates, Inc., 2015

2015
[12]

J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei. ImageNet: A Large-Scale Hierarchical Image Database. In CVPR, 2009. 13

2009
[13]

Adversarial feature learning

Jeff Donahue, Philipp Kr ¨ahenb¨uhl, and Trevor Darrell. Adversarial feature learning. arXiv preprint arXiv:1605.09782 , 2016

2016 arXiv
[14]

Training generative neural networks via maximum mean discrepancy optimization

Gintare Karolina Dziugaite, Daniel M Roy, and Zoubin Ghahramani. Training generative neural networks via maximum mean discrepancy optimization. UAI, 2015

2015
[15]

Z. Feng, J. Lai, and X. Xie. Learning view-speciﬁc deep networks for person re-identiﬁcation. IEEE Transactions on Image Processing , PP(99):1–1, 2018

2018
[16]

Yuan Gao, Jiayi Ma, and Alan L. Yuille. Semi-supervised sparse representation based classiﬁcation for face recognition with insufﬁcient labeled samples. CoRR, abs/1609.03279, 2016

2016 arXiv
[17]

Generative adversarial nets

Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In NIPS, 2014

2014
[18]

Goodfellow

Ian J. Goodfellow. NIPS 2016 tutorial: Generative adversarial networks. CoRR, abs/1701.00160, 2017

2016 arXiv
[19]

A kernel two-sample test

Arthur Gretton, Karsten M Borgwardt, Malte J Rasch, Bernhard Sch¨olkopf, and Alexander Smola. A kernel two-sample test. Journal of Machine Learning Research , 13(Mar):723–773, 2012

2012
[20]

Digital selection and analogue ampliﬁ- cation coexist in a cortex-inspired silicon circuit

Richard HR Hahnloser, Rahul Sarpeshkar, Misha A Mahowald, Rodney J Douglas, and H Sebastian Seung. Digital selection and analogue ampliﬁ- cation coexist in a cortex-inspired silicon circuit. Nature, 405(6789):947, 2000

2000
[21]

Deep residual learning for image recognition

Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In CVPR, 2016

2016
[22]

Nlh: A blind pixel-level non-local method for real-world image denoising, 2019

Yingkun Hou, Jun Xu, Guanghai Liu, Li Liu, Fan Zhu, and Ling Shao. Nlh: A blind pixel-level non-local method for real-world image denoising, 2019

2019
[23]

Hospedales, and Jakob Verbeek

Guosheng Hu, Xiaojiang Peng, Yongxin Yang, Timothy M. Hospedales, and Jakob Verbeek. Frankenstein: Learning deep face representations using small data. IEEE Trans. Image Processing, 27(1):293–303, 2018

2018
[24]

Huang, Manu Ramesh, Tamara Berg, and Erik Learned-Miller

Gary B. Huang, Manu Ramesh, Tamara Berg, and Erik Learned-Miller. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. Technical Report 07-49, 2007

2007
[25]

Batch normalization: Accelerating deep network training by reducing internal covariate shift

Sergey Ioffe and Christian Szegedy. Batch normalization: Accelerating deep network training by reducing internal covariate shift. CoRR, abs/1502.03167, 2015

2015 arXiv
[26]

Semi-supervised learning with deep generative models

Diederik P Kingma, Shakir Mohamed, Danilo Jimenez Rezende, and Max Welling. Semi-supervised learning with deep generative models. In Advances in Neural Information Processing Systems , pages 3581– 3589, 2014

2014
[27]

Suspect identiﬁcation based on descriptive facial attributes

Brendan F Klare, Scott Klum, Joshua C Klontz, Emma Taborsky, Tayfun Akgul, and Anil K Jain. Suspect identiﬁcation based on descriptive facial attributes. In IJCB, 2014

2014
[28]

Learning multiple layers of features from tiny images

Alex Krizhevsky. Learning multiple layers of features from tiny images. 2009

2009
[29]

Imagenet classiﬁcation with deep convolutional neural networks

Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classiﬁcation with deep convolutional neural networks. In NIPS, 2012

2012
[30]

Facetracer: A search engine for large collections of images with faces

Neeraj Kumar, Peter Belhumeur, and Shree Nayar. Facetracer: A search engine for large collections of images with faces. In ECCV, 2008

2008
[31]

Describable visual attributes for face veriﬁcation and image search

Neeraj Kumar, Alexander Berg, Peter N Belhumeur, and Shree Nayar. Describable visual attributes for face veriﬁcation and image search. IEEE Transactions on Pattern Analysis and Machine Intelligence , 33(10):1962–1977, 2011

1962
[32]

Attribute and simile classiﬁers for face veriﬁcation

Neeraj Kumar, Alexander C Berg, Peter N Belhumeur, and Shree K Nayar. Attribute and simile classiﬁers for face veriﬁcation. In ICCV, 2009

2009
[33]

Gradient-based learning applied to document recognition

Yann LeCun, L ´eon Bottou, Yoshua Bengio, and Patrick Haffner. Gradient-based learning applied to document recognition. Proceedings of the IEEE , 86(11):2278–2324, 1998

1998
[34]

Mmd gan: Towards deeper understanding of moment matching network

Chun-Liang Li, Wei-Cheng Chang, Yu Cheng, Yiming Yang, and Barnabas Poczos. Mmd gan: Towards deeper understanding of moment matching network. In I. Guyon, U. V . Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information...

2017
[35]

J. Li, B. Li, J. Xu, R. Xiong, and W. Gao. Fully connected network- based intra prediction for image coding. IEEE Transactions on Image Processing, PP(99):1–1, 2018

2018
[36]

A hybrid l1-l0 layer decomposition model for tone mapping

Zhetong Liang, Jun Xu, David Zhang, Zisheng Cao, and Lei Zhang. A hybrid l1-l0 layer decomposition model for tone mapping. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , June 2018

2018
[37]

Classiﬁcation and regression by randomforest

Andy Liaw, Matthew Wiener, et al. Classiﬁcation and regression by randomforest. R news, 2(3):18–22, 2002

2002
[38]

D. Liu, Z. Wang, Y . Fan, X. Liu, Z. Wang, S. Chang, X. Wang, and T. S. Huang. Learning temporal dynamics for video super-resolution: A deep learning approach. IEEE Transactions on Image Processing , PP(99):1–1, 2018

2018
[39]

Liu and J

N. Liu and J. Han. A deep spatial contextual long-term recurrent convolutional network for saliency detection. IEEE Transactions on Image Processing, PP(99):1–1, 2018

2018
[40]

Deep learning face attributes in the wild

Ziwei Liu, Ping Luo, Xiaogang Wang, and Xiaoou Tang. Deep learning face attributes in the wild. In ICCV, 2015

2015
[41]

Auxiliary deep generative models

Lars Maaløe, Casper Kaae Sønderby, Søren Kaae Sønderby, and Ole Winther. Auxiliary deep generative models. arXiv preprint arXiv:1602.05473, 2016

2016 arXiv
[42]

Maas, Awni Y

Andrew L. Maas, Awni Y . Hannun, and Andrew Y . Ng. Rectiﬁer nonlinearities improve neural network acoustic models. 2013

2013
[43]

Distributional smoothing with virtual adversarial training

Takeru Miyato, Shin-ichi Maeda, Masanori Koyama, Ken Nakae, and Shin Ishii. Distributional smoothing with virtual adversarial training. arXiv preprint arXiv:1507.00677 , 2015

2015 arXiv
[44]

Nascimento and Gustavo Carneiro

Jacinto C. Nascimento and Gustavo Carneiro. Deep learning on sparse manifolds for faster object segmentation. IEEE Trans. Image Processing, 26(10):4978–4990, 2017

2017
[45]

Reading digits in natural images with unsupervised feature learning

Yuval Netzer, Tao Wang, Adam Coates, Alessandro Bissacco, Bo Wu, and Andrew Y Ng. Reading digits in natural images with unsupervised feature learning. In ICCV workshop, 2011

2011
[46]

Learning to generate reviews and discovering sentiment

Alec Radford, Rafal J ´ozefowicz, and Ilya Sutskever. Learning to generate reviews and discovering sentiment. CoRR, abs/1704.01444, 2017

2017 arXiv
[47]

Unsupervised representation learning with deep convolutional generative adversarial networks

Alec Radford, Luke Metz, and Soumith Chintala. Unsupervised representation learning with deep convolutional generative adversarial networks. In ICLR, 2016

2016
[48]

Semi-supervised learning with ladder networks

Antti Rasmus, Mathias Berglund, Mikko Honkala, Harri Valpola, and Tapani Raiko. Semi-supervised learning with ladder networks. In NIPS, 2015

2015
[49]

Semi- supervised self-training of object detection models

Chuck Rosenberg, Martial Hebert, and Henry Schneiderman. Semi- supervised self-training of object detection models. 2005

2005
[50]

Moon: A mixed objective optimization network for the recognition of facial attributes

Ethan M Rudd, Manuel G ¨unther, and Terrance E Boult. Moon: A mixed objective optimization network for the recognition of facial attributes. In ECCV, 2016

2016
[51]

Berg, and Li Fei-Fei

Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, Alexander C. Berg, and Li Fei-Fei. ImageNet Large Scale Visual Recognition Challenge. IJCV, 115(3):211–252, 2015

2015
[52]

Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Radford, and Xi Chen

Tim Salimans, Ian J. Goodfellow, Wojciech Zaremba, Vicki Cheung, Alec Radford, and Xi Chen. Improved techniques for training gans. In NIPS, 2016

2016
[53]

Simonyan and A

K. Simonyan and A. Zisserman. Very deep convolutional networks for large-scale image recognition. In ICLR, 2015

2015
[54]

Unsupervised and semi-supervised learning with categorical generative adversarial networks

Jost Tobias Springenberg. Unsupervised and semi-supervised learning with categorical generative adversarial networks. 2016

2016
[55]

Least squares support vector machine classiﬁers

Johan AK Suykens and Joos Vandewalle. Least squares support vector machine classiﬁers. Neural processing letters , 9(3):293–300, 1999

1999
[56]

Going deeper with convolutions

Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, and Andrew Rabinovich. Going deeper with convolutions. In CVPR, 2015

2015
[57]

Learning multi- instance deep discriminative patterns for image classiﬁcation

Peng Tang, Xinggang Wang, Bin Feng, and Wenyu Liu. Learning multi- instance deep discriminative patterns for image classiﬁcation. IEEE Trans. Image Processing, 26(7):3385–3396, 2017

2017
[58]

Video salient object de- tection via fully convolutional networks

Wenguan Wang, Jianbing Shen, and Ling Shao. Video salient object de- tection via fully convolutional networks. IEEE Trans. Image Processing, 27(1):38–49, 2018

2018
[59]

Ranet: Ranking attention network for fast video object segmentation

Ziqin Wang, Jun Xu, Li Liu, Fan Zhu, and Ling Shao. Ranet: Ranking attention network for fast video object segmentation. In The IEEE International Conference on Computer Vision (ICCV) , Oct 2019

2019
[60]

Empirical evaluation of rectiﬁed activations in convolutional network

Bing Xu, Naiyan Wang, Tianqi Chen, and Mu Li. Empirical evaluation of rectiﬁed activations in convolutional network. CoRR, abs/1505.00853, 2015

2015 arXiv
[61]

J. Xu. Nonlocal self-similarity based prior modeling for image denois- ing. PhD thesis, The Hong Kong Polytechnic University, 2018

2018
[62]

J. Xu, Y . Huang, L. Liu, F. Zhu, X. Hou, and L. Shao. Noisy-as-clean: Learning unsupervised denoising from the corrupted image, 2019

2019
[63]

J. Xu, H. Li, Z. Liang, D. Zhang, and L. Zhang. Real-world noisy image denoising: A new benchmark. arXiv:1804.02603, 2018

2018 arXiv
[64]

J. Xu, D. Ren, L. Zhang, and D. Zhang. Patch group based bayesian learning for blind image denoising. Asian Conference on Computer Vision Workshop, pages 79–95, 2016

2016
[65]

J. Xu, L. Zhang, and D. Zhang. External prior guided internal prior learning for real-world noisy image denoising. IEEE Transactions on Image Processing, 27(6):2996–3010, June 2018

2018
[66]

J. Xu, L. Zhang, and D. Zhang. A trilateral weighted sparse coding scheme for real-world image denoising. In ECCV, 2018. 14

2018
[67]

J. Xu, L. Zhang, D. Zhang, and X. Feng. Multi-channel weighted nuclear norm minimization for real color image denoising. In ICCV, 2017

2017
[68]

J. Xu, L. Zhang, W. Zuo, D. Zhang, and X. Feng. Patch group based nonlocal self-similarity prior learning for image denoising. In ICCV, pages 244–252, 2015

2015
[69]

Sparse, collabora- tive, or nonnegative representation: Which helps pattern classiﬁcation? Pattern Recognition, 88:679 – 688, 2019

Jun Xu, Wangpeng An, Lei Zhang, and David Zhang. Sparse, collabora- tive, or nonnegative representation: Which helps pattern classiﬁcation? Pattern Recognition, 88:679 – 688, 2019

2019
[70]

Reweighted sparse subspace clustering

Jun Xu, Kui Xu, Ke Chen, and Jishou Ruan. Reweighted sparse subspace clustering. Computer Vision and Image Understanding , 138(0):25–37, 2015

2015
[71]

Star: A structure and texture aware retinex model, 2019

Jun Xu, Mengyang Yu, Li Liu, Fan Zhu, Dongwei Ren, Yingkun Hou, Haoqian Wang, and Ling Shao. Star: A structure and texture aware retinex model, 2019

2019
[72]

Tstss: A two-stage training subset selection framework for cross version defect prediction

Zhou Xu, Shuai Li, Xiapu Luo, Jin Liu, Tao Zhang, Yutian Tang, Jun Xu, Peipei Yuan, and Jacky Keung. Tstss: A two-stage training subset selection framework for cross version defect prediction. Journal of Systems and Software , 2019

2019
[73]

Cross version defect prediction with representative data via sparse subset selection

Zhou Xu, Shuai Li, Yutian Tang, Xiapu Luo, Tao Zhang, Jin Liu, and Jun Xu. Cross version defect prediction with representative data via sparse subset selection. In Proceedings of the 26th Conference on Program Comprehension, ICPC ’18, pages 132–143. ACM, 2018

2018
[74]

Panda: Pose aligned networks for deep attribute modeling

Ning Zhang, Manohar Paluri, Marc’Aurelio Ranzato, Trevor Darrell, and Lubomir Bourdev. Panda: Pose aligned networks for deep attribute modeling. In CVPR, 2014

2014
[75]

newdna-prot: Prediction of dna-binding proteins by employing support vector machine and a comprehensive sequence representation

Yanping Zhang, Jun Xu, Wei Zheng, Chen Zhang, Xingye Qiu, Ke Chen, and Jishou Ruan. newdna-prot: Prediction of dna-binding proteins by employing support vector machine and a comprehensive sequence representation. Computational Biology and Chemistry , 52(0):51–59, 2014

2014
[76]

Energy-based generative adversarial network

Junbo Jake Zhao, Micha ¨el Mathieu, and Yann LeCun. Energy-based generative adversarial network. CoRR, abs/1609.03126, 2016

2016 arXiv
[77]

Semi-supervised regression with co- training

Zhi-Hua Zhou and Ming Li. Semi-supervised regression with co- training. In IJCAI, volume 5, pages 908–913, 2005

2005
[78]

Introduction to semi-supervised learning

Xiaojin Zhu and Andrew B Goldberg. Introduction to semi-supervised learning. Synthesis lectures on artiﬁcial intelligence and machine learning, 3:1–130, 2009

2009
[79]

Zhuang, Z

L. Zhuang, Z. Zhou, S. Gao, J. Yin, Z. Lin, and Y . Ma. Label information guided graph construction for semi-supervised learning. IEEE Transactions on Image Processing , 26(9):4182–4192, Sept 2017

2017

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

A pid controller approach for stochastic optimization of deep networks

Wangpeng An, Haoqian Wang, Qingyun Sun, Jun Xu, Qionghai Dai, and Lei Zhang. A pid controller approach for stochastic optimization of deep networks. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018

work page 2018

[2] [2]

Towards principled methods for training generative adversarial networks

Martin Arjovsky and L ´eon Bottou. Towards principled methods for training generative adversarial networks. In ICLR, 2017

work page 2017

[3] [3]

Wasserstein generative adversarial networks

Mart ´ın Arjovsky, Soumith Chintala, and L ´eon Bottou. Wasserstein generative adversarial networks. In Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW, Australia, 6-11 August 2017 , pages 214–223, 2017

work page 2017

[4] [4]

Poof: Part-based one-vs.-one features for ﬁne-grained categorization, face veriﬁcation, and attribute estimation

Thomas Berg and Peter Belhumeur. Poof: Part-based one-vs.-one features for ﬁne-grained categorization, face veriﬁcation, and attribute estimation. In CVPR, 2013

work page 2013

[5] [5]

Describing people: A poselet-based approach to attribute classiﬁcation

Lubomir Bourdev, Subhransu Maji, and Jitendra Malik. Describing people: A poselet-based approach to attribute classiﬁcation. In ICCV, 2011

work page 2011

[6] [6]

Semi- Supervised Learning

Olivier Chapelle, Bernhard Schlkopf, and Alexander Zien. Semi- Supervised Learning. The MIT Press, 1st edition, 2010

work page 2010

[7] [7]

Chen and W

K. Chen and W. Tao. Convolutional regression for visual tracking. IEEE Transactions on Image Processing , PP(99):1–1, 2018

work page 2018

[8] [8]

Goodfellow, and Jonathon Shlens

Tianqi Chen, Ian J. Goodfellow, and Jonathon Shlens. Net2net: Accel- erating learning via knowledge transfer. In ICLR, 2016

work page 2016