MLP-Mixer

MLP-Mixer: An all-MLP Architecture for Vision

摘要

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.

使用方式

Predict imageUse the modelTest Command

from mmpretrain import inference_model

predict = inference_model('mlp-mixer-base-p16_3rdparty_64xb64_in1k', 'demo/bird.JPEG')
print(predict['pred_class'])
print(predict['pred_score'])

Models and results

Image Classification on ImageNet-1k

模型	预训练	Params (M)	Flops (G)	Top-1 (%)	Top-5 (%)	配置文件	下载
mlp-mixer-base-p16_3rdparty_64xb64_in1k*	从头训练	59.88	12.61	76.68	92.25	config	model
mlp-mixer-large-p16_3rdparty_64xb64_in1k*	从头训练	208.20	44.57	72.34	88.02	config	model

Models with * are converted from the timm. The config files of these models are only for inference. We haven’t reproduce the training results.

引用

@misc{tolstikhin2021mlpmixer,
      title={MLP-Mixer: An all-MLP Architecture for Vision},
      author={Ilya Tolstikhin and Neil Houlsby and Alexander Kolesnikov and Lucas Beyer and Xiaohua Zhai and Thomas Unterthiner and Jessica Yung and Andreas Steiner and Daniel Keysers and Jakob Uszkoreit and Mario Lucic and Alexey Dosovitskiy},
      year={2021},
      eprint={2105.01601},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}