AtteConDA: Attention-Based Conflict Suppression in Multi-Condition Diffusion Models and Synthetic Data Augmentation

TL;DR

This paper introduces AtteConDA, a multi-condition diffusion model with conflict suppression to generate structurally faithful images for autonomous driving, enhancing data augmentation and high-level scene understanding.

Contribution

It proposes a novel conflict modeling approach for multi-condition image generation, improving structural preservation in synthetic data for autonomous driving tasks.

Findings

The conflict-aware model enables better structural fidelity in generated images.

The framework improves data augmentation for high-level driving tasks.

Evaluation shows enhanced scene consistency compared to baseline methods.

Abstract

Recent conditional image generation methods can improve controllability by generating images that are faithful to conditions such as sketches, human poses, segmentation maps, and depth. By applying these techniques to image augmentation while preserving annotations, generated images can be used as additional training data and can improve recognition performance. However, for high-level driving tasks such as traffic-rule extraction and driving-behavior understanding, simply using annotations as conditions is insufficient. Instead, images must be augmented while preserving the detailed high-level structure of the original scene. One possible solution is to use multiple conditions so that generated images retain diverse structural cues after generation. However, when multiple conditions are used, conflicts among conditions can prevent reliable structure preservation. In this work, we input semantic segmentation, depth, and edges extracted from the original image into a multi-condition image generation model, thereby providing rich structural information as conditions. We further propose a modeling approach for handling conflicts among multiple conditions and show that it enables image generation with stronger structural preservation. We also build a generation framework and evaluation protocol for driving tasks, establishing a basis for comparison with prior and future models. As a result, this work contributes to image generation research by addressing condition conflicts in multi-condition generation and provides an important step toward mitigating data scarcity in high-level autonomous-driving tasks.