Publications | Datasets | CV

---

GOPRO_Large dataset

GOPRO_Large dataset is proposed for dynamic scene deblurring. Training and Test set are publicly available.

Google Drive

• GOPRO_Large: Blurry and sharp image pairs. Blurry images includes both gamma corrected and not corrected (linear CRF) versions.
• GOPRO_Large_all: All the sharp images used to generate blurry images. You can generate new blurry images by accumulating differing number of sharp frames.

SNU CVLab Server

• GOPRO_Large: Blurry and sharp image pairs. Blurry images includes both gamma corrected and not corrected (linear CRF) versions.
• GOPRO_Large_all: All the sharp images used to generate blurry images. You can generate new blurry images by accumulating differing number of sharp frames.

Reference

The GOPRO_Large dataset was proposed in CVPR 2017. If you find our dataset useful for your research, please consider citing our work:

• paper
• supplementary
• slides
• video

@InProceedings{Nah_2017_CVPR,
  author = {Nah, Seungjun and Kim, Tae Hyun and Lee, Kyoung Mu},
  title = {Deep Multi-Scale Convolutional Neural Network for Dynamic Scene Deblurring},
  booktitle = {CVPR},
  month = {July},
  year = {2017}
}

Github repository

The source code for training deblurring models are available:

• Torch7: DeepDeblur_release
• PyTorch: DeepDeblur-PyTorch

LICENSE

GOPRO dataset is released under CC BY 4.0 license.