MearcAll papers

arXiv 1511.06434 · cs.LG cs.CV

Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

Alec Radford, Luke Metz, Soumith Chintala

Claims were extracted automatically by reducer-1 on 2026-11-12. The paper's license does not allow republishing it, so long claims are shortened to a quotation. Read the full text on arXiv.

No reproduction recordedclm_100d3c1b
We introduce a class of CNNs called deep convolutional generative adversarial networks (DCGANs), that have certain architectural constraints, and demonstrate that they are a strong candidate for unsupervised learning.

ABSTRACT · page 1 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_90f7c60b
We propose that one way to build good image representations is by training Generative Adversarial Networks (GANs) (Goodfellow et al., 2014), and later reusing parts of the generator and discriminator … (excerpt)

I NTRODUCTION · page 1 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_00099701
1 arXiv:1511.06434v2 [cs.LG] 7 Jan 2016 Under review as a conference paper at ICLR 2016 • We show that the generators have interesting vector arithmetic properties allowing for easy manipulation … (excerpt)

I NTRODUCTION · page 1 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_633507de
We show samples from one epoch of training (Fig.2), mimicking online learning, in addition to samples after convergence (Fig.3), as an opportunity to demonstrate that our model is not producing … (excerpt)

LSUN · page 4 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_76e28a8f
When using a very large amount of feature maps (4800) this technique achieves 80.6% accuracy.

C LASSIFYING CIFAR-10 USING GAN S AS A FEATURE EXTRACTOR · page 6 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_499b7713
This achieves 82.8% accuracy, out performing all K-means based approaches.

C LASSIFYING CIFAR-10 USING GAN S AS A FEATURE EXTRACTOR · page 6 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedres_fa6ff9b0
It achieves a signficantly higher 28.87% validation error.

C LASSIFYING CIFAR-10 USING GAN S AS A FEATURE EXTRACTOR · page 6 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_9daff68d
We demonstrate that an unsupervised DCGAN trained on a large image dataset can also learn a hierarchy of features that are interesting.

V ISUALIZING THE DISCRIMINATOR FEATURES · page 7 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_4ec797ed
In addition to the object manipulation shown in (Fig. 7), we demonstrate that face pose is also modeled linearly inZ space (Fig. 8).

V ECTOR ARITHMETIC ON FACE SAMPLES · page 8 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_7e578cfc
We propose a more stable set of architectures for training generative adversarial networks and we give evidence that adversarial networks learn good representations of images for supervised learning and generative … (excerpt)

C ONCLUSION AND FUTURE WORK · page 9 · Automated extraction, not reviewed · 0 reviewed reproductions

No reproduction recordedclm_543d8ec9
We propose to apply standard classification metrics to a conditional version of our model, evaluating the conditional distributions learned.

E VALUATING DCGAN S CAPABILITY TO CAPTURE DATA DISTRIBUTIONS · page 14 · Automated extraction, not reviewed · 0 reviewed reproductions

Reproduced one of these claims?

Record your result in the Mearc app on your own machine, then use Publish this result with this paper's id (1511.06434) and the claim id shown next to the claim. Mearc creates a file that holds only what you choose to publish. Add it to the public claim data by pull request; a maintainer other than you reviews it before it counts.