B. Choudhury, R. Swanson, F. Heide, G. Wetzstein, W. Heidrich
Convolutional sparse coding, High-level vision tasks, Low-level image reconstruction
Convolutional sparse coding (CSC) is a promising direction
for unsupervised learning in computer vision. In contrast
to recent supervised methods, CSC allows for convolutional
image representations to be learned that are equally
useful for high-level vision tasks and low-level image reconstruction
and can be applied to a wide range of tasks without
problem-specific retraining. Due to their extreme memory
requirements, however, existing CSC solvers have so far
been limited to low-dimensional problems and datasets using
a handful of low-resolution example images at a time.
In this paper, we propose a new approach to solving CSC
as a consensus optimization problem, which lifts these limitations.
By learning CSC features from large-scale image
datasets for the first time, we achieve significant quality
improvements in a number of imaging tasks. Moreover,
the proposed method enables new applications in highdimensional
feature learning that has been intractable using
existing CSC methods. This is demonstrated for a variety
of reconstruction problems across diverse problem
domains, including 3D multispectral demosaicing and 4D
light field view synthesis.