Neural Differential Equations for Single Image Super-resolution
This work addresses a practical problem in image processing for researchers and practitioners, but it is incremental as it benchmarks existing methods on a new task.
The authors tackled the challenge of applying Neural Differential Equations to single image super-resolution, finding that discrete sensitivity analysis outperforms the adjoint method in stability and achieving performance matching a state-of-the-art super-resolution model.
Although Neural Differential Equations have shown promise on toy problems such as MNIST, they have yet to be successfully applied to more challenging tasks. Inspired by variational methods for image restoration relying on partial differential equations, we choose to benchmark several forms of Neural DEs and backpropagation methods on single image super-resolution. The adjoint method previously proposed for gradient estimation has no theoretical stability guarantees; we find a practical case where this makes it unusable, and show that discrete sensitivity analysis has better stability. In our experiments, differential models match the performance of a state-of-the art super-resolution model.