Deblurring is the task of removing motion blurs that usually occur in photos shot with hand-held cameras, when there are moving objects in the scene. Blurs not only reduce the human perception about the quality of the image, but also complicate computer vision analyses.
This tutorial demonstrates Single Image Motion Deblurring with DeblurGAN-v2 in OpenVINO, by first converting the VITA-Group/DeblurGANv2 model to OpenVINO Intermediate Representation (OpenVINO IR) format. For more information about the model, see the deblurgan-v2 model documentation.
For more information, refer to the following research paper:
Kupyn, O., Martyniuk, T., Wu, J., & Wang, Z. (2019). Deblurgan-v2: Deblurring (orders-of-magnitude) faster and better. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 8878-8887).
The image above is a sample of the test subset of the GOPRO dataset. This dataset is one of the three that were used to train DeblurGAN-v2, and contains pairs of blurred and sharp images.
The authors release the dataset under the CC BY 4.0 license.
If you have not installed all required dependencies, follow the Installation Guide.