MLP-Mixer: An all-MLP Architecture for Vision

This seems to be quite an important recent contribution to advancing computer vision! It maybe a game changer!

 E.g. Andrew Ng emphasizes this paper in his latest newsletter: "Why use a complex model when a simple one will do? New work shows that the simplest multilayer neural network, with a small twist, can perform some tasks as well as today’s most sophisticated architectures."!

Here is the abstract of this paper:

"Convolutional Neural Networks (CNNs) are the go-to model for computer vision. Recently, attention-based networks, such as the Vision Transformer, have also become popular. In this paper we show that while convolutions and attention are both sufficient for good performance, neither of them are necessary. We present MLP-Mixer, an architecture based exclusively on multi-layer perceptrons (MLPs). MLP-Mixer contains two types of layers: one with MLPs applied independently to image patches (i.e. "mixing" the per-location features), and one with MLPs applied across patches (i.e. "mixing" spatial information). When trained on large datasets, or with modern regularization schemes, MLP-Mixer attains competitive scores on image classification benchmarks, with pre-training and inference cost comparable to state-of-the-art models. We hope that these results spark further research beyond the realms of well established CNNs and Transformers."

[2105.01601] MLP-Mixer: An all-MLP Architecture for Vision