Pivotal Auto-Encoder via Self-Normalizing ReLU
This addresses a limitation for real-world applications of auto-encoders where noise is unpredictable, though it appears incremental as it builds on existing sparse auto-encoder frameworks.
The paper tackles the problem of sparse auto-encoders degrading in performance when test-time noise differs from training noise, proposing a method that yields models invariant to noise levels, with experimental results showing significant improvement in stability against varying noise types.
Sparse auto-encoders are useful for extracting low-dimensional representations from high-dimensional data. However, their performance degrades sharply when the input noise at test time differs from the noise employed during training. This limitation hinders the applicability of auto-encoders in real-world scenarios where the level of noise in the input is unpredictable. In this paper, we formalize single hidden layer sparse auto-encoders as a transform learning problem. Leveraging the transform modeling interpretation, we propose an optimization problem that leads to a predictive model invariant to the noise level at test time. In other words, the same pre-trained model is able to generalize to different noise levels. The proposed optimization algorithm, derived from the square root lasso, is translated into a new, computationally efficient auto-encoding architecture. After proving that our new method is invariant to the noise level, we evaluate our approach by training networks using the proposed architecture for denoising tasks. Our experimental results demonstrate that the trained models yield a significant improvement in stability against varying types of noise compared to commonly used architectures.