FE-Wireframe Dataset

✏️ Github   |   📑 Paper    |   🖼️ Viewer

This is the FE-Wireframe dataset, designed for motion-blurred image line segment detection with events.

Summary

The FE-Wireframe dataset is constructed to provide synthetic data for motion-blur scenarios without requiring manual line annotations. Starting from the labeled Wireframe dataset, we employ the ESIM simulator to generate event-based synthetic sequences. The overall data generation pipeline is shown as follows.

• File structure:

|-- events_raw
    |-- 0000001.npz
    |-- 0000002.npz
    |-- ...
|-- images-blur
    |-- 0000001.png
    |-- 0000002.png
    |-- ...
|-- images-start
    |-- 0000001.png
    |-- 0000002.png
    |-- ...
|-- images-end
    |-- 0000001.png
    |-- 0000002.png
    |-- ...
|-- train.jsonl
|-- test.jsonl

• Number of samples:
  • Train: 5,000
  • Test: 462

Download

• Download with huggingface-hub

python3 -m pip install huggingface-hub
huggingface-cli download --repo-type dataset lh9171338/FE-Wireframe --local-dir ./

• Download with Git

git lfs install
git clone https://huggingface.co/datasets/lh9171338/FE-Wireframe

Usage

• Load the dataset from Hugging Face Hub

from datasets import load_dataset

ds = load_dataset("lh9171338/FE-Wireframe", trust_remote_code=True)
print(ds)
# DatasetDict({
#     train: Dataset({
#         features: ['blur_image', 'start_image', 'end_image', 'events', 'H', 'image_size', 'junc', 'flow', 'lines', 'edges_positive'],
#         num_rows: 5000
#     })
#     test: Dataset({
#         features: ['blur_image', 'start_image', 'end_image', 'events', 'H', 'image_size', 'junc', 'flow', 'lines', 'edges_positive'],
#         num_rows: 462
#     })
# })

• Load the dataset from local

from datasets import load_dataset

ds = load_dataset("FE-Wireframe", trust_remote_code=True)
print(ds)
# DatasetDict({
#     train: Dataset({
#         features: ['blur_image', 'start_image', 'end_image', 'events', 'H', 'image_size', 'junc', 'flow', 'lines', 'edges_positive'],
#         num_rows: 5000
#     })
#     test: Dataset({
#         features: ['blur_image', 'start_image', 'end_image', 'events', 'H', 'image_size', 'junc', 'flow', 'lines', 'edges_positive'],
#         num_rows: 462
#     })
# })

• Load the dataset with jsonl files

import jsonlines

with jsonlines.open("test.jsonl") as reader:
    infos = list(reader)
print(infos[0].keys())
# dict_keys(['filename', 'image_size', 'H', 'junc', 'flow', 'lines', 'edges_positive'])

Citation

@ARTICLE{10323537,
  author={Yu, Huai and Li, Hao and Yang, Wen and Yu, Lei and Xia, Gui-Song},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence}, 
  title={Detecting Line Segments in Motion-Blurred Images With Events}, 
  year={2023},
  pages={1-16},
  doi={10.1109/TPAMI.2023.3334877}
}