# Edge Guided GANs with Contrastive Learning for Semantic Image Synthesis

arXiv: https://arxiv.org/abs/2003.13898 · HTML: https://mlsanity.com/paper/2003.13898/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Mar 2020.

## Built on by (most cited first, 8 of 8)

- [Graph Transformer GANs for Graph-Constrained House Generation](https://mlsanity.com/paper/2303.08225/) (2023)
- [SCP-Diff: Spatial-Categorical Joint Prior for Diffusion Based Semantic Image Synthesis](https://mlsanity.com/paper/2403.09638/) (2024)
- [Graph Transformer GANs with Graph Masked Modeling for Architectural Layout Generation](https://mlsanity.com/paper/2401.07721/) (2024)
- [3D CoCa v2: Contrastive Learners with Test-Time Search for Generalizable Spatial Intelligence](https://mlsanity.com/paper/2601.06496/) (2026)
- [Boosting Few-Shot Detection with Large Language Models and Layout-to-Image Synthesis](https://mlsanity.com/paper/2410.06841/) (2024)
- [Hierarchical Cross-Attention Network for Virtual Try-On](https://mlsanity.com/paper/2411.15542/) (2024)
- [Data-driven Camera and Lidar Simulation Models for Autonomous Driving: A Review from Generative Models to Volume Renderers](https://mlsanity.com/paper/2402.10079/) (2024)
- [Introducing Resizable Region Packing Problem in Image Generation, with a Heuristic Solution](https://mlsanity.com/paper/2509.16363/) (2025)

## Abstract (arXiv)

We propose a novel ECGAN for the challenging semantic image synthesis task. Although considerable improvement has been achieved, the quality of synthesized images is far from satisfactory due to three largely unresolved challenges. 1) The semantic labels do not provide detailed structural information, making it difficult to synthesize local details and structures. 2) The widely adopted CNN operations such as convolution, down-sampling, and normalization usually cause spatial resolution loss and thus cannot fully preserve the original semantic information, leading to semantically inconsistent results. 3) Existing semantic image synthesis methods focus on modeling local semantic information from a single input semantic layout. However, they ignore global semantic information of multiple input semantic layouts, i.e., semantic cross-relations between pixels across different input layouts. To tackle 1), we propose to use edge as an intermediate representation which is further adopted to guide image generation via a proposed attention guided edge transfer module. Edge information is produced by a convolutional generator and introduces detailed structure information. To tackle 2), we design an effective module to selectively highlight class-dependent feature maps according to the original semantic layout to preserve the semantic information. To tackle 3), inspired by current methods in contrastive learning, we propose a novel contrastive learning method, which aims to enforce pixel embeddings belonging to the same semantic class to generate more similar image content than those from different classes. Doing so can capture more semantic relations by explicitly exploring the structures of labeled pixels from multiple input semantic layouts. Experiments on three challenging datasets show that our ECGAN achieves significantly better results than state-of-the-art methods.