Rethinking Text-to-Image as Semantic-Aware Data Augmentation for Indoor Scene Recognition

2026-06-16

AI Summary

This work addresses the performance limitations of indoor scene image recognition caused by illumination variations, occlusions, and complex layouts, compounded by the scarcity of real annotated data. To mitigate these challenges, the authors propose a semantic-aware data augmentation approach based on Stable Diffusion that generates semantically consistent synthetic indoor images to enhance model generalization. Additionally, they introduce a lightweight detection mechanism grounded in Diffusion Reconstruction Error (DIRE) to reliably identify synthetic images and prevent potential misuse. Experimental results demonstrate that the proposed method significantly improves the accuracy of recognition models such as MobileNetV3 on the MIT Indoor Scenes dataset, while achieving 100% detection precision for images generated by Stable Diffusion.

Abstract

In the realm of computer vision, indoor image recognition presents challenges due to the intricate interplay of lighting conditions, occlusions, and diverse object arrangements within confined spaces. To address the lacks of training indoor images, we introduce a novel approach leveraging Stable Diffusion (SD) for the generation of synthetic images, which serve as a powerful data augmentation tool. The utilization of SD offers a principled framework for synthesizing diverse and realistic indoor scenes, thereby enriching the training data pool for robust indoor image recognition models. Experimental findings on the MIT Indoor Scene dataset reveal the potential of our proposed approach in enhancing the training of deep models when authentic data is limited. Furthermore, to prevent the misuse of SD synthetic images, we introduce a counter measure based on DIffusion Reconstruction Error (DIRE). The powerful DIRE presentation enables training robust classifiers only using lightweight deep models. Experiments show that our approach can perfectly recognize SD generated images with the accuracy of 100% using MobilenetV3.

Problem

• indoor scene recognition
• data scarcity
• synthetic image generation
• computer vision
• data augmentation

Innovation

• Semantic-Aware Data Augmentation
• Stable Diffusion
• Indoor Scene Recognition
• Diffusion Reconstruction Error (DIRE)
• Synthetic Image Detection