Can I read Tunable Convolutions with Parametric Multi-Loss Optimization on EtoBox?

Tunable Convolutions with Parametric Multi-Loss Optimization by Maggioni, Matteo; Tanay, Thomas; Babiloni, Francesca; McDonagh, Steven; Leonardis, Aleš is a scholarly article available to read on EtoBox.

What is Tunable Convolutions with Parametric Multi-Loss Optimization about?

Behavior of neural networks is irremediably determined by the specific loss and data used during training. However it is often desirable to tune the model at inference time based on external factors such as preferences of the user or dynamic characteristics of the data. This is especially important to balance the perception-distortion trade-off of ill-posed image-to-image translation tasks. In this work, we propose to optimize a parametric tunable convolutional layer, which includes a number of different kernels, using a parametric multi-loss, which includes an equal number of objectives. Our key insight is to use a shared set of parameters to dynamically interpolate both the objectives and the kernels. During training, these parameters are sampled at random to explicitly optimize all possible combinations of objectives and consequently disentangle their effect into the corresponding kernels. During inference, these parameters become interactive inputs of the model hence enabling reliable and consistent control over the model behavior. Extensive experimental results demonstrate that our tunable convolutions effectively work as a drop-in replacement for traditional convolutions in e

Author

Maggioni, Matteo; Tanay, Thomas; Babiloni, Francesca; McDonagh, Steven; Leonardis, Aleš

Published

2023

Language

EN