Methods › General › Loss Functions › GAN Least Squares Loss

GAN Least Squares Loss

421 papers tagged archive 2025-07-28

Introduced by Xudong Mao et al. in Least Squares Generative Adversarial Networks

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

GAN Least Squares Loss is a least squares loss function for generative adversarial networks. Minimizing this objective function is equivalent to minimizing the Pearson χ² divergence. The objective function (here for LSGAN) can be defined as:

min_DV_(LS)(D) = 1/2𝔼_(𝐱 ∼p_(data)(𝐱))[(D(𝐱) - b)²] + 1/2𝔼_(𝐳∼p_(data)(𝐳))[(D(G(𝐳)) - a)²]

min_GV_(LS)(G) = 1/2𝔼_(𝐳 ∼p_𝐳(𝐳))[(D(G(𝐳)) - c)²]

where a and b are the labels for fake data and real data and c denotes the value that G wants D to believe for fake data.

PaperSourceSee Code · eriklindernoren/PyTorch-GAN

Papers archive 2025-07-28

30 shown of 421, 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 271 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
Translation104
Image-to-Image Translation89
Image Generation51
Generative Adversarial Network47
Domain Adaptation38
Data Augmentation32
Semantic Segmentation29
Style Transfer28
Segmentation25
SSIM24
Diagnostic16
Computed Tomography (CT)15
Denoising14
Image Segmentation13
Object Detection13
Super-Resolution13
object-detection13
Anatomy12
Diversity12
Unsupervised Domain Adaptation12

Usage over time archive 2025-07-28

Papers per year tagged with GAN Least Squares Loss: 2016 to 2025, peak 79 79 0 2016: 1 paper 2016 2017: 13 papers 2017 2018: 33 papers 2018 2019: 51 papers 2019 2020: 79 papers 2020 2021: 58 papers 2021 2022: 59 papers 2022 2023: 54 papers 2023 2024: 50 papers 2024 2025: 23 papers 2025
Papers per year the archive tags with this method, by the paper's archive date (421 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

Loss Functions

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