Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Sep 6, 2025

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Published on: March 18, 2019

Label-Guided Generative Adversarial Network for Realistic Image Synthesis.

Junchen Zhu, Lianli Gao, Jingkuan Song

IEEE Transactions on Pattern Analysis and Machine Intelligence

|June 28, 2022

Summary

We introduce Lab2Pix, a framework for generating realistic images from labels. It uses novel techniques like Double-Guided Normalization and Label Guided Spatial Co-Attention to improve image quality and efficiency.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Accelerated Exploration of Empty Material Compositional Space: Mg-Fe-B Ternary Metal Borides.

Journal of the American Chemical Society·2024

Same author

Admission albumin-globulin ratio associated with delayed cerebral ischemia following aneurysmal subarachnoid hemorrhage.

Frontiers in neurology·2024

Same author

Computational discovery of two-dimensional tetragonal group IV-V monolayers.

RSC advances·2024

Same author

Nickel-Catalyzed Direct Fluorosulfonylation of Vinyl Bromides and Benzyl Bromides for Sulfonyl Fluorides.

Organic letters·2024

Same author

Preoperative Prediction of Occult Level V Lymph Node Metastasis in Papillary Thyroid Carcinoma: Development and Validation of a Radiomics-Driven Nomogram Model.

Academic radiology·2024

Same author

Self-guided Knowledge-Injected Graph Neural Network for Alzheimer's Diseases.

Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention·2024

Area of Science:

Computer Vision
Artificial Intelligence
Machine Learning

Background:

Generating photo-realistic images from sparse labels is a significant challenge in image-to-image translation.
Existing methods struggle with the large domain gap between simple labels and complex, detailed images.

Purpose of the Study:

To propose a general framework, Lab2Pix, for generating high-quality images from various labels.
To address challenges in information extraction from labels and bridging the label-image domain gap.

Main Methods:

Developed Double-Guided Normalization (DG-Norm) for semantic guidance and differentiation within normalization layers.
Introduced Label Guided Spatial Co-Attention (LSCA) for efficient incremental visual information learning.
Utilized Hierarchical Perceptual Discriminators with Foreground Enhancement Masks and a sharp enhancement loss for realistic and sharp image generation.

More Related Videos

Photorealistic Learned Landscapes for Augmented Reality

Photorealistic Learned Landscapes for Augmented Reality

Published on: June 27, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Related Experiment Videos

Last Updated: Sep 6, 2025

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Generating Strictly Controlled Stimuli for Figure Recognition Experiments

Published on: March 18, 2019

Photorealistic Learned Landscapes for Augmented Reality

Photorealistic Learned Landscapes for Augmented Reality

Published on: June 27, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Main Results:

Lab2Pix significantly outperforms state-of-the-art methods in label-to-image generation tasks.
The framework demonstrates strong quantitative and qualitative results in both unpaired (Lab2Pix-V1) and paired (Lab2Pix-V2) settings.
Extensive experiments on various datasets validate the effectiveness of the proposed methods.

Conclusions:

Lab2Pix provides an effective general framework for label-to-image synthesis.
The proposed DG-Norm and LSCA mechanisms successfully tackle the challenges of information extraction and domain gap bridging.
The method achieves superior performance, paving the way for more advanced image generation techniques.