datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
stride-architecture-components-v1
STRIDE Architecture Threat Modeling Dataset (AWS & Azure)
📌 Overview
This dataset was created to enable automatic STRIDE threat modeling from cloud architecture diagrams (AWS and Azure).
The goal is to detect architectural components in diagrams and support automated threat identification based on data flows and trust boundaries.
Annotations were created using Label Studio in YOLO format.
Total images: 4190Total classes: 32
🎯 Purpose
Detect cloud… See the full description on the dataset page: https://huggingface.co/datasets/guillherms/stride-architecture-components-v1.component-static-buildspowerline-components-and-faults
Powerline Components and Faults Dataset
Overview
The Powerline Components and Faults Dataset is a dataset designed for object detection tasks involving powerline components and associated faults. It provides images of powerline infrastructure along with annotated bounding boxes for various components and faults.
This dataset is augmented with mosaic augmentation useful for training and evaluating models on powerline inspection, maintenance, and safety applications.… See the full description on the dataset page: https://huggingface.co/datasets/docmhvr/powerline-components-and-faults.powerline-components-and-faults
Powerline Components and Faults Dataset
Overview
The Powerline Components and Faults Dataset is a dataset designed for object detection tasks involving powerline components and associated faults. It provides images of powerline infrastructure along with annotated bounding boxes for various components and faults.
This dataset is augmented with mosaic augmentation useful for training and evaluating models on powerline inspection, maintenance, and safety… See the full description on the dataset page: https://huggingface.co/datasets/meekL/powerline-components-and-faults.mirrai-sd-training-open-components
MirrAI SD Training Open Components
This dataset repo contains the redistributable subset that was actually used in the MirrAI Stable Diffusion hair inpainting training pipeline.
Included In This Release
face_sketches_refined_generation/
images_512/: 139 processed 512px training images
masks/: hair, face protect, and cloth protect masks
controls/canny/: canny control inputs
face_crops/: identity-preserving face crops
manifests/recon_hair_external.jsonl: training manifest… See the full description on the dataset page: https://huggingface.co/datasets/siik/mirrai-sd-training-open-components.electrical-componentsflowchat_components基于
galirage/FC-Detection
shreyanshu09/Block_Diagram
的重新标注。
使用makesense.ai对数据集的重新标注，标注分为以下几类
0: text
1: arrow
2: terminator
3: data
4: process
5: decision
6: connection
7: arrow_start
8: arrow_end
此分类参考论文 Arrow-Guided VLM: Enhancing Flowchart Understanding via Arrow Direction Encoding，也对论文所提到的数据集galirage/FC-Detection进行了重新标注，并加入了手动标注的几十张shreyanshu09/Block_Diagram流程图
示例：
powerline-components-and-faults
Powerline Components and Faults Dataset
Overview
The Powerline Components and Faults Dataset is a dataset designed for object detection tasks involving powerline components and associated faults. It provides images of powerline infrastructure along with annotated bounding boxes for various components and faults.
This dataset is augmented with mosaic augmentation useful for training and evaluating models on powerline inspection, maintenance, and safety applications.… See the full description on the dataset page: https://huggingface.co/datasets/cathyhy/powerline-components-and-faults.TW-Kai_components_256jedi-components-imageschange-components-bahia