FFDNet: Toward a Fast and Flexible Solution for CNN based Image Denoising
This provides a more practical and efficient solution for image denoising applications, though it is incremental as it builds on existing CNN-based methods.
The paper tackled the problem of image denoising by addressing the limitations of existing discriminative methods that require multiple models for different noise levels and lack flexibility for spatially variant noise, resulting in a single network (FFDNet) that handles noise levels from 0 to 75, removes spatially variant noise, and achieves faster speed than BM3D on CPU without performance loss.
Due to the fast inference and good performance, discriminative learning methods have been widely studied in image denoising. However, these methods mostly learn a specific model for each noise level, and require multiple models for denoising images with different noise levels. They also lack flexibility to deal with spatially variant noise, limiting their applications in practical denoising. To address these issues, we present a fast and flexible denoising convolutional neural network, namely FFDNet, with a tunable noise level map as the input. The proposed FFDNet works on downsampled sub-images, achieving a good trade-off between inference speed and denoising performance. In contrast to the existing discriminative denoisers, FFDNet enjoys several desirable properties, including (i) the ability to handle a wide range of noise levels (i.e., [0, 75]) effectively with a single network, (ii) the ability to remove spatially variant noise by specifying a non-uniform noise level map, and (iii) faster speed than benchmark BM3D even on CPU without sacrificing denoising performance. Extensive experiments on synthetic and real noisy images are conducted to evaluate FFDNet in comparison with state-of-the-art denoisers. The results show that FFDNet is effective and efficient, making it highly attractive for practical denoising applications.