Object-Centric Dataset Resources for Constrained-Data Image Generation and Augmentation

2026-06-19

AI Summary

This work addresses the scarcity of standardized, object-centric low-resource datasets for data augmentation and image generation in domains such as pedestrian analysis in smart cities and traffic sign detection. To bridge this gap, the authors systematically construct and publicly release three object-centered datasets—Cityscapes-Pedestrian, TrafficSigns, and COCO PottedPlant—each providing 256×256 cropped images, bounding box annotations, and contextual information that collectively capture representative challenges including dense occlusion, high contrast, and diverse backgrounds. Through standardized cropping, privacy-preserving blurring, reproducible subset splits, and a reconstruction protocol, the study delivers over 12,000 high-quality annotated samples, enabling controllable image generation and robust model evaluation under low-data regimes.

Abstract

Object-centric image generation is important in settings with few labeled examples, including pedestrian analysis in smart-city scenes, traffic-sign inspection, and domain-specific object detection. Synthetic images are most useful for training and evaluation when datasets preserve object structure, bounding boxes, visual diversity, and realistic context. Existing image datasets usually target classification, detection, or scene understanding rather than controlled object-centric generation and augmentation with limited class-specific data. We present a shareable collection of three object-centric dataset resources: Cityscapes-Pedestrian, TrafficSigns, and COCO PottedPlant. The collection standardizes 256-by-256 object-centric crops and bounding-box annotations across three regimes: dense pedestrian scenes with privacy blur and occlusion, cleaner high-contrast traffic signs, and context-diverse potted-plant scenes. The release contains 3,009 TrafficSigns samples, 2,156 Cityscapes-Pedestrian manifest records, and 7,679 COCO PottedPlant manifest records. The larger COCO-derived manifest preserves contextual and multi-instance diversity, while equal-size subsets can be drawn with a fixed random seed for controlled comparisons. The release provides direct TrafficSigns data where redistribution is permitted, together with scripts, manifests, box-level annotation tables, checksums, and reconstruction documentation for the Cityscapes- and COCO-derived subsets. It is available through the Latzi/object-centric-low-data-datasets GitHub repository and Zenodo DOI 10.5281/zenodo.20573001. The collection supports label and split inspection, subset creation, reconstruction from upstream data, and evaluation of object-centric image generation or synthetic-data augmentation methods on shared records.

Problem

• object-centric generation
• low-data image generation
• synthetic data augmentation
• constrained-data setting
• dataset resource

Innovation

• object-centric generation
• low-data image augmentation
• synthetic data
• standardized dataset
• controlled subset sampling