Global Adaptive Filtering Layer for Computer Vision

Abstract

We devise a universal adaptive neural layer to "learn" optimal frequency filter for each image together with the weights of the base neural network that performs some computer vision task. The proposed approach takes the source image in the spatial domain, automatically selects the best frequencies from the frequency domain, and transmits the inverse-transform image to the main neural network. Remarkably, such a simple add-on layer dramatically improves the performance of the main network regardless of its design. We observe that the light networks gain a noticeable boost in the performance metrics; whereas, the training of the heavy ones converges faster when our adaptive layer is allowed to "learn" alongside the main architecture. We validate the idea in four classical computer vision tasks: classification, segmentation, denoising, and erasing, considering popular natural and medical data benchmarks.

The manuscript is under consideration at Computer Vision and Image Understanding. 28 pages, 25 figures (main article and supplementary material). V.S. and I.B contributed equally, D.V.D is Corresponding author

References in corpus (5)

• Attention U-Net: Learning Where to Look for the Pancreas
• Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
• The Real-World-Weight Cross-Entropy Loss Function: Modeling the Costs of Mislabeling
• A Review of Convolutional Neural Networks for Inverse Problems in Imaging
• Comparison of Image Quality Models for Optimization of Image Processing Systems