</> Code with Papers by Syntology

Methods › Computer Vision › Generative Models › Pix2Pix

Pix2Pix

132 papers tagged archive 2025-07-28

archive 2025-07-28 Description, source and code snippet are the archive's method entry.

Pix2Pix is a conditional image-to-image translation architecture that uses a conditional GAN objective combined with a reconstruction loss. The conditional GAN objective for observed images x, output images y and the random noise vector z is:

ℒ_(cGAN)(G, D) =𝔼_(x,y)[logD(x, y)]+ 𝔼_(x,z)[log(1 − D(x, G(x, z))]

We augment this with a reconstruction term:

ℒ_(L1)(G) = 𝔼_(x,y,z)[||y - G(x, z)||₁]

and we get the final objective as:

G^* = min_Gmax_Dℒ_(cGAN)(G, D) + λℒ_(L1)(G)

The architectures employed for the generator and discriminator closely follow DCGAN, with a few modifications:

Source: Image-to-Image Translation with Conditional Adversarial NetworksSee Code · junyanz/pytorch-CycleGAN-and-pix2pix

Papers archive 2025-07-28

30 shown of 132, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.

Tasks archive 2025-07-28

20 shown of 133 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.

TaskPapers
Translation41
Image-to-Image Translation37
Generative Adversarial Network31
Image Generation19
SSIM14
Semantic Segmentation12
Segmentation11
Data Augmentation9
Diagnostic7
Colorization6
Style Transfer6
Autonomous Driving5
Anatomy4
Decoder4
Deep Learning4
Denoising4
Object4
Object Detection4
object-detection4
Autonomous Vehicles3

Usage over time archive 2025-07-28

Papers per year tagged with Pix2Pix: 2016 to 2025, peak 28 28 0 2016: 1 paper 2016 2017: 2 papers 2017 2018: 1 paper 2018 2019: 12 papers 2019 2020: 14 papers 2020 2021: 28 papers 2021 2022: 15 papers 2022 2023: 23 papers 2023 2024: 28 papers 2024 2025: 8 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (132 dated). Bars are counts, not a trend claim.

Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).

Categories archive 2025-07-28

Generative Models

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections