More than 1 year has passed since last update.


Last updated at Posted at 2021-07-01



2021 Q2


Paper MLP-Mixer: An all-MLP Architecture for Vision
Date 4 May 2021
Code https://github.com/google-research/vision_transformer
Authors Ilya Tolstikhin, Neil Houlsby, Alexander Kolesnikov, Lucas Beyer, Xiaohua Zhai, Thomas Unterthiner, Jessica Yung, Andreas Steiner, Daniel Keysers, Jakob Uszkoreit, Mario Lucic, Alexey Dosovitskiy
Org. Google Research, Brain Team
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.


Paper Do You Even Need Attention? A Stack of Feed-Forward Layers Does Surprisingly Well on ImageNet
Date 6 May 2021
Code https://github.com/lukemelas/do-you-even-need-attention
Authors Luke Melas-Kyriazi
Org. Oxford University
The strong performance of vision transformers on image classification and other vision tasks is often attributed to the design of their multi-head attention layers. However, the extent to which attention is responsible for this strong performance remains unclear. In this short report, we ask: is the attention layer even necessary? Specifically, we replace the attention layer in a vision transformer with a feed-forward layer applied over the patch dimension. The resulting architecture is simply a series of feed-forward layers applied over the patch and feature dimensions in an alternating fashion. In experiments on ImageNet, this architecture performs surprisingly well: a ViT/DeiT-base-sized model obtains 74.9% top-1 accuracy, compared to 77.9% and 79.9% for ViT and DeiT respectively. These results indicate that aspects of vision transformers other than attention, such as the patch embedding, may be more responsible for their strong performance than previously thought. We hope these results prompt the community to spend more time trying to understand why our current models are as effective as they are.


Paper ResMLP: Feedforward networks for image classification with data-efficient training
Date Fri, 7 May 2021
Code https://github.com/facebookresearch/deit
Authors Hugo Touvron, Piotr Bojanowski, Mathilde Caron, Matthieu Cord, Alaaeldin El-Nouby, Edouard Grave, Gautier Izacard, Armand Joulin, Gabriel Synnaeve, Jakob Verbeek, Hervé Jégou
Org. Facebook AI, Sorbonne University, Inria, valeo.ai
We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We also train ResMLP models in a self-supervised setup, to further remove priors from employing a labelled dataset. Finally, by adapting our model to machine translation we achieve surprisingly good results.We share pre-trained models and our code based on the Timm library.


Paper Pay Attention to MLPs
Date 17 May 2021
Authors Hanxiao Liu, Zihang Dai, David R. So, Quoc V. Le
Org. Google Research, Brain Team
Transformers have become one of the most important architectural innovations in deep learning and have enabled many breakthroughs over the past few years. Here we propose a simple network architecture, gMLP, based on MLPs with gating, and show that it can perform as well as Transformers in key language and vision applications. Our comparisons show that self-attention is not critical for Vision Transformers, as gMLP can achieve the same accuracy. For BERT, our model achieves parity with Transformers on pretraining perplexity and is better on some downstream NLP tasks. On finetuning tasks where gMLP performs worse, making the gMLP model substantially larger can close the gap with Transformers. In general, our experiments show that gMLP can scale as well as Transformers over increased data and compute.


Paper S2-MLP: Spatial-Shift MLP Architecture for Vision
Date 14 Jun 2021
Authors Tan Yu, Xu Li, Yunfeng Cai, Mingming Sun, Ping Li
Org. Baidu Research
Recently, visual Transformer (ViT) and its following works abandon the convolution and exploit the self-attention operation, attaining a comparable or even higher accuracy than CNNs. More recently, MLP-Mixer abandons both the convolution and the self-attention operation, proposing an architecture containing only MLP layers. To achieve cross-patch communications, it devises an additional token-mixing MLP besides the channel-mixing MLP. It achieves promising results when training on an extremely large-scale dataset. But it cannot achieve as outstanding performance as its CNN and ViT counterparts when training on medium-scale datasets such as ImageNet1K and ImageNet21K. The performance drop of MLP-Mixer motivates us to rethink the token-mixing MLP. We discover that the token-mixing MLP is a variant of the depthwise convolution with a global reception field and spatial-specific configuration. But the global reception field and the spatial-specific property make token-mixing MLP prone to over-fitting. In this paper, we propose a novel pure MLP architecture, spatial-shift MLP (S2-MLP). Different from MLP-Mixer, our S2-MLP only contains channel-mixing MLP. We utilize a spatial-shift operation for communications between patches. It has a local reception field and is spatial-agnostic. It is parameter-free and efficient for computation. The proposed S2-MLP attains higher recognition accuracy than MLP-Mixer when training on ImageNet-1K dataset. Meanwhile, S2-MLP accomplishes as excellent performance as ViT on ImageNet-1K dataset with considerably simpler architecture and fewer FLOPs and parameters.


Paper Vision Permutator: A Permutable MLP-Like Architecture for Visual Recognition
Date 23 Jun 2021
Code https://github.com/Andrew-Qibin/VisionPermutator
Authors Qibin Hou, Zihang Jiang, Li Yuan, Ming-Ming Cheng, Shuicheng Yan, Jiashi Feng
Org. National University of Singapore / Nankai University / Sea AI Labs / ECE, NUS & Sea AI Labsg
In this paper, we present Vision Permutator, a conceptually simple and data efficient MLP-like architecture for visual recognition. By realizing the importance of the positional information carried by 2D feature representations, unlike recent MLP-like models that encode the spatial information along the flattened spatial dimensions, Vision Permutator separately encodes the feature representations along the height and width dimensions with linear projections. This allows Vision Permutator to capture long-range dependencies along one spatial direction and meanwhile preserve precise positional information along the other direction. The resulting position-sensitive outputs are then aggregated in a mutually complementing manner to form expressive representations of the objects of interest. We show that our Vision Permutators are formidable competitors to convolutional neural networks (CNNs) and vision transformers. Without the dependence on spatial convolutions or attention mechanisms, Vision Permutator achieves 81.5% top-1 accuracy on ImageNet without extra large-scale training data (e.g., ImageNet-22k) using only 25M learnable parameters, which is much better than most CNNs and vision transformers under the same model size constraint. When scaling up to 88M, it attains 83.2% top-1 accuracy. We hope this work could encourage research on rethinking the way of encoding spatial information and facilitate the development of MLP-like models.

CCS token-mixing MLP

Paper Rethinking Token-Mixing MLP for MLP-based Vision Backbone
Date 28 Jun 2021
Authors Tan Yu, Xu Li, Yunfeng Cai, Mingming Sun, Ping Li
Org. Cognitive Computing Lab Baidu Research
In the past decade, we have witnessed rapid progress in the machine vision backbone. By introducing the inductive bias from the image processing, convolution neural network (CNN) has achieved excellent performance in numerous computer vision tasks and has been established as \emph{de facto} backbone. In recent years, inspired by the great success achieved by Transformer in NLP tasks, vision Transformer models emerge. Using much less inductive bias, they have achieved promising performance in computer vision tasks compared with their CNN counterparts. More recently, researchers investigate using the pure-MLP architecture to build the vision backbone to further reduce the inductive bias, achieving good performance. The pure-MLP backbone is built upon channel-mixing MLPs to fuse the channels and token-mixing MLPs for communications between patches. In this paper, we re-think the design of the token-mixing MLP. We discover that token-mixing MLPs in existing MLP-based backbones are spatial-specific, and thus it is sensitive to spatial translation. Meanwhile, the channel-agnostic property of the existing token-mixing MLPs limits their capability in mixing tokens. To overcome those limitations, we propose an improved structure termed as Circulant Channel-Specific (CCS) token-mixing MLP, which is spatial-invariant and channel-specific. It takes fewer parameters but achieves higher classification accuracy on ImageNet1K benchmark.

2021 Q3


Paper What Makes for Hierarchical Vision Transformer?
Date 5 Jul 2021
Authors Yuxin Fang, Xinggang Wang, Rui Wu, Wenyu Liu
Org. Huazhong University of Science & Technology , Horizon Robotics
Recent studies indicate that hierarchical Vision Transformer with a macro architecture of interleaved non-overlapped window-based self-attention & shifted-window operation is able to achieve state-of-the-art performance in various visual recognition tasks, and challenges the ubiquitous convolutional neural networks (CNNs) using densely slid kernels. Most follow-up works attempt to replace the shifted-window operation with other kinds of cross-window communication paradigms, while treating self-attention as the de-facto standard for window-based information aggregation. In this manuscript, we question whether self-attention is the only choice for hierarchical Vision Transformer to attain strong performance, and the effects of different kinds of cross-window communication. To this end, we replace self-attention layers with embarrassingly simple linear mapping layers, and the resulting proof-of-concept architecture termed as LinMapper can achieve very strong performance in ImageNet-1k image recognition. Moreover, we find that LinMapper is able to better leverage the pre-trained representations from image recognition and demonstrates excellent transfer learning properties on downstream dense prediction tasks such as object detection and instance segmentation. We also experiment with other alternatives to self-attention for content aggregation inside each non-overlapped window under different cross-window communication approaches, which all give similar competitive results. Our study reveals that the \textbf{macro architecture} of Swin model families, other than specific aggregation layers or specific means of cross-window communication, may be more responsible for its strong performance and is the real challenger to the ubiquitous CNN's dense sliding window paradigm. Code and models will be publicly available to facilitate future research.


Paper AS-MLP: An Axial Shifted MLP Architecture for Vision
Date 18 Jul 2021
Code https://github.com/svip-lab/AS-MLP
Authors Dongze Lian, Zehao Yu, Xing Sun, Shenghua Gao
Org. ShanghaiTech University, Tencent
An Axial Shifted MLP architecture (AS-MLP) is proposed in this paper. Different from MLP-Mixer, where the global spatial feature is encoded for the information flow through matrix transposition and one token-mixing MLP, we pay more attention to the local features communication. By axially shifting channels of the feature map, AS-MLP is able to obtain the information flow from different axial directions, which captures the local dependencies. Such an operation enables us to utilize a pure MLP architecture to achieve the same local receptive field as CNN-like architecture. We can also design the receptive field size and dilation of blocks of AS-MLP, etc, just like designing those of convolution kernels. With the proposed AS-MLP architecture, our model obtains 83.3% Top-1 accuracy with 88M parameters and 15.2 GFLOPs on the ImageNet-1K dataset. Such a simple yet effective architecture outperforms all MLP-based architectures and achieves competitive performance compared to the transformer-based architectures (e.g., Swin Transformer) even with slightly lower FLOPs. In addition, AS-MLP is also the first MLP-based architecture to be applied to the downstream tasks (e.g., object detection and semantic segmentation). The experimental results are also impressive. Our proposed AS-MLP obtains 51.5 mAP on the COCO validation set and 49.5 MS mIoU on the ADE20K dataset, which is competitive compared to the transformer-based architectures.


Paper CycleMLP: A MLP-like Architecture for Dense Prediction
Date 21 Jul 2021
Code https://github.com/ShoufaChen/CycleMLP
Authors Shoufa Chen, Enze Xie, Chongjian Ge, Ding Liang, Ping Luo
Org. The University of Hong Kong , SenseTime Research
This paper presents a simple MLP-like architecture, CycleMLP, which is a versatile backbone for visual recognition and dense predictions, unlike modern MLP architectures, e.g., MLP-Mixer, ResMLP, and gMLP, whose architectures are correlated to image size and thus are infeasible in object detection and segmentation. CycleMLP has two advantages compared to modern approaches. (1) It can cope with various image sizes. (2) It achieves linear computational complexity to image size by using local windows. In contrast, previous MLPs have quadratic computations because of their fully spatial connections. We build a family of models that surpass existing MLPs and achieve a comparable accuracy (83.2%) on ImageNet-1K classification compared to the state-of-the-art Transformer such as Swin Transformer (83.3%) but using fewer parameters and FLOPs. We expand the MLP-like models' applicability, making them a versatile backbone for dense prediction tasks. CycleMLP aims to provide a competitive baseline on object detection, instance segmentation, and semantic segmentation for MLP models. In particular, CycleMLP achieves 45.1 mIoU on ADE20K val, comparable to Swin (45.2 mIOU)


Paper S2-MLPv2: Improved Spatial-Shift MLP Architecture for Vision
Date 2 Aug 2021
Authors Tan Yu, Xu Li, Yunfeng Cai, Mingming Sun, Ping Li
Org. Baidu Research
Recently, MLP-based vision backbones emerge. MLP-based vision architectures with less inductive bias achieve competitive performance in image recognition compared with CNNs and vision Transformers. Among them, spatial-shift MLP (S2-MLP), adopting the straightforward spatial-shift operation, achieves better performance than the pioneering works including MLP-mixer and ResMLP. More recently, using smaller patches with a pyramid structure, Vision Permutator (ViP) and Global Filter Network (GFNet) achieve better performance than S2-MLP.In this paper, we improve the S2-MLP vision backbone. We expand the feature map along the channel dimension and split the expanded feature map into several parts. We conduct different spatial-shift operations on split parts.Meanwhile, we exploit the split-attention operation to fuse these split parts. Moreover, like the counterparts, we adopt smaller-scale patches and use a pyramid structure for boosting the image recognition accuracy. We term the improved spatial-shift MLP vision backbone as S2-MLPv2. Using 55M parameters, our medium-scale model, S2-MLPv2-Medium achieves an 83.6% top-1 accuracy on the ImageNet-1K benchmark using 224×224 images without self-attention and external training data.


Paper RaftMLP: Do MLP-based Models Dream of Winning Over Computer Vision?
Date 9 Aug 2021
Code https://github.com/okojoalg/raft-mlp
Authors Yuki Tatsunami, Masato Taki
Org. Rikkyo University
For the past ten years, CNN has reigned supreme in the world of computer vision, but recently, Transformer is on the rise. However, the quadratic computational cost of self-attention has become a severe problem of practice. There has been much research on architectures without CNN and self-attention in this context. In particular, MLP-Mixer is a simple idea designed using MLPs and hit an accuracy comparable to the Vision Transformer. However, the only inductive bias in this architecture is the embedding of tokens. Thus, there is still a possibility to build a non-convolutional inductive bias into the architecture itself, and we built in an inductive bias using two simple ideas. A way is to divide the token-mixing block vertically and horizontally. Another way is to make spatial correlations denser among some channels of token-mixing. With this approach, we were able to improve the accuracy of the MLP-Mixer while reducing its parameters and computational complexity. Compared to other MLP-based models, the proposed model, named RaftMLP has a good balance of computational complexity, the number of parameters, and actual memory usage. In addition, our work indicates that MLP-based models have the potential to replace CNNs by adopting inductive bias.


Paper Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Date 17 Aug 2021
Code https://github.com/microsoft/Swin-Transformer
Authors Ze Liu, Yutong Lin, Yue Cao, Han Hu, Yixuan Wei, Zheng Zhang, Stephen Lin, Baining Guo
Org. Microsoft Research Asia
This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains, such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. To address these differences, we propose a hierarchical Transformer whose representation is computed with \textbf{S}hifted \textbf{win}dows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation (53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and +2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones. The hierarchical design and the shifted window approach also prove beneficial for all-MLP architectures.


Paper Hire-MLP: Vision MLP via Hierarchical Rearrangement
Date 30 Aug 2021
Code https://github.com/ggjy/Hire-Wave-MLP.pytorch
Authors Jianyuan Guo, Yehui Tang, Kai Han, Xinghao Chen, Han Wu, Chao Xu, Chang Xu, Yunhe Wang
Org. Huawei Technologies, Peking University, University of Sydney
This paper presents Hire-MLP, a simple yet competitive vision MLP architecture via hierarchical rearrangement. Previous vision MLPs like MLP-Mixer are not flexible for various image sizes and are inefficient to capture spatial information by flattening the tokens. Hire-MLP innovates the existing MLP-based models by proposing the idea of hierarchical rearrangement to aggregate the local and global spatial information while being versatile for downstream tasks. Specifically, the inner-region rearrangement is designed to capture local information inside a spatial region. Moreover, to enable information communication between different regions and capture global context, the cross-region rearrangement is proposed to circularly shift all tokens along spatial directions. The proposed Hire-MLP architecture is built with simple channel-mixing MLPs and rearrangement operations, thus enjoys high flexibility and inference speed. Experiments show that our Hire-MLP achieves state-of-the-art performance on the ImageNet-1K benchmark. In particular, Hire-MLP achieves an 83.4% top-1 accuracy on ImageNet, which surpasses previous Transformer-based and MLP-based models with better trade-off for accuracy and throughput.


Paper Sparse-MLP: A Fully-MLP Architecture with Conditional Computation
Date 5 Sep 2021
Authors Yuxuan Lou, Fuzhao Xue, Zangwei Zheng, Yang You
Org. National University of Singapore
Mixture of Experts (MoE) with sparse conditional computation has been proved an effective architecture for scaling attention-based models to more parameters with comparable computation cost. In this paper, we propose Sparse-MLP, scaling the recent MLP-Mixer model with sparse MoE layers, to achieve a more computation-efficient architecture. We replace a subset of dense MLP blocks in the MLP-Mixer model with Sparse blocks. In each Sparse block, we apply two stages of MoE layers: one with MLP experts mixing information within channels along image patch dimension, one with MLP experts mixing information within patches along the channel dimension. Besides, to reduce computational cost in routing and improve experts capacity, we design Re-represent layers in each Sparse block. These layers are to re-scale image representations by two simple but effective linear transformations. By pre-training on ImageNet-1k with MoCo v3 algorithm, our models can outperform dense MLP models with comparable parameters and less computational cost on several downstream image classification tasks.


Paper ConvMLP: Hierarchical Convolutional MLPs for Vision
Date 9 Sep 2021
Code https://github.com/SHI-Labs/Convolutional-MLPs
Authors Jiachen Li, Ali Hassani, Steven Walton, Humphrey Shi
Org. University of Oregon, UIUC, Picsart AI Research
MLP-based architectures, which consist of a sequence of consecutive multi-layer perceptron blocks, have recently been found to reach comparable results to convolutional and transformer-based methods. However, most adopt spatial MLPs which take fixed dimension inputs, therefore making it difficult to apply them to downstream tasks, such as object detection and semantic segmentation. Moreover, single-stage designs further limit performance in other computer vision tasks and fully connected layers bear heavy computation. To tackle these problems, we propose ConvMLP: a hierarchical Convolutional MLP for visual recognition, which is a light-weight, stage-wise, co-design of convolution layers, and MLPs. In particular, ConvMLP-S achieves 76.8% top-1 accuracy on ImageNet-1k with 9M parameters and 2.4G MACs (15% and 19% of MLP-Mixer-B/16, respectively). Experiments on object detection and semantic segmentation further show that visual representation learned by ConvMLP can be seamlessly transferred and achieve competitive results with fewer parameters.


Paper Sparse MLP for Image Recognition: Is Self-Attention Really Necessary?
Date 12 Sep 2021
Authors Chuanxin Tang, Yucheng Zhao, Guangting Wang, Chong Luo, Wenxuan Xie, Wenjun Zeng
Org. Microsoft Research Asia / University of Science and Technology of China
Transformers have sprung up in the field of computer vision. In this work, we explore whether the core self-attention module in Transformer is the key to achieving excellent performance in image recognition. To this end, we build an attention-free network called sMLPNet based on the existing MLP-based vision models. Specifically, we replace the MLP module in the token-mixing step with a novel sparse MLP (sMLP) module. For 2D image tokens, sMLP applies 1D MLP along the axial directions and the parameters are shared among rows or columns. By sparse connection and weight sharing, sMLP module significantly reduces the number of model parameters and computational complexity, avoiding the common over-fitting problem that plagues the performance of MLP-like models. When only trained on the ImageNet-1K dataset, the proposed sMLPNet achieves 81.9% top-1 accuracy with only 24M parameters, which is much better than most CNNs and vision Transformers under the same model size constraint. When scaling up to 66M parameters, sMLPNet achieves 83.4% top-1 accuracy, which is on par with the state-of-the-art Swin Transformer. The success of sMLPNet suggests that the self-attention mechanism is not necessarily a silver bullet in computer vision.

2021 Q4


Paper Convolutional Gated MLP: Combining Convolutions & gMLP
Date 6 Nov 2021
Authors A.Rajagopal, V. Nirmala
Org. Indian Institute of Technology, Queen Mary’s College
To the best of our knowledge, this is the first paper to introduce Convolutions to Gated MultiLayer Perceptron and contributes an implementation of this novel Deep Learning architecture. Google Brain introduced the gMLP in May 2021. Microsoft introduced Convolutions in Vision Transformer in Mar 2021. Inspired by both gMLP and CvT, we introduce convolutional layers in gMLP. CvT combined the power of Convolutions and Attention. Our implementation combines the best of Convolutional learning along with spatial gated MLP. Further, the paper visualizes how CgMLP learns. Visualizations show how CgMLP learns from features such as outline of a car. While Attention was the basis of much of recent progress in Deep Learning, gMLP proposed an approach that doesn't use Attention computation. In Transformer based approaches, a whole lot of Attention matrixes need to be learnt using vast amount of training data. In gMLP, the fine tunning for new tasks can be challenging by transfer learning with smaller datasets. We implement CgMLP and compares it with gMLP on CIFAR dataset. Experimental results explore the power of generaliza-tion of CgMLP, while gMLP tend to drastically overfit the training data.To summarize, the paper contributes a novel Deep Learning architecture and demonstrates the learning mechanism of CgMLP through visualizations, for the first time in literature.


Paper An Image Patch is a Wave: Phase-Aware Vision MLP
Date 24 Nov 2021
Code https://github.com/ggjy/Hire-Wave-MLP.pytorch
Authors Yehui Tang, Kai Han, Jianyuan Guo, Chang Xu, Yanxi Li, Chao Xu, Yunhe Wang
Org. Peking University / Huawei Noah’s Ark Lab / University of Sydney
Different from traditional convolutional neural network (CNN) and vision transformer, the multilayer perceptron (MLP) is a new kind of vision model with extremely simple architecture that only stacked by fully-connected layers. An input image of vision MLP is usually split into multiple tokens (patches), while the existing MLP models directly aggregate them with fixed weights, neglecting the varying semantic information of tokens from different images. To dynamically aggregate tokens, we propose to represent each token as a wave function with two parts, amplitude and phase. Amplitude is the original feature and the phase term is a complex value changing according to the semantic contents of input images. Introducing the phase term can dynamically modulate the relationship between tokens and fixed weights in MLP. Based on the wave-like token representation, we establish a novel Wave-MLP architecture for vision tasks. Extensive experiments demonstrate that the proposed Wave-MLP is superior to the state-of-the-art MLP architectures on various vision tasks such as image classification, object detection and semantic segmentation.


Paper MorphMLP: A Self-Attention Free, MLP-Like Backbone for Image and Video
Date 24 Nov 2021
Authors David Junhao Zhang, Kunchang Li, Yunpeng Chen, Yali Wang, Shashwat Chandra, Yu Qiao, Luoqi Liu, Mike Zheng Shou
Org. National University of Singapore / Meitu, Inc / Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences / University of Chinese Academy of Sciences / Shanghai AI Laboratory
Self-attention has become an integral component of the recent network architectures, e.g., Transformer, that dominate major image and video benchmarks. This is because self-attention can flexibly model long-range information. For the same reason, researchers make attempts recently to revive Multiple Layer Perceptron (MLP) and propose a few MLP-Like architectures, showing great potential. However, the current MLP-Like architectures are not good at capturing local details and lack progressive understanding of core details in the images and/or videos. To overcome this issue, we propose a novel MorphMLP architecture that focuses on capturing local details at the low-level layers, while gradually changing to focus on long-term modeling at the high-level layers. Specifically, we design a Fully-Connected-Like layer, dubbed as MorphFC, of two morphable filters that gradually grow its receptive field along the height and width dimension. More interestingly, we propose to flexibly adapt our MorphFC layer in the video domain. To our best knowledge, we are the first to create a MLP-Like backbone for learning video representation. Finally, we conduct extensive experiments on image classification, semantic segmentation and video classification. Our MorphMLP, such a self-attention free backbone, can be as powerful as and even outperform self-attention based models.


Paper RepMLPNet: Hierarchical Vision MLP with Re-parameterized Locality
Date 21 Dec 2021
Code https://github.com/DingXiaoH/RepMLP
Authors Xiaohan Ding, Honghao Chen, Xiangyu Zhang, Jungong Han, Guiguang Ding
Org. Tsinghua University / Chinese Academy of Sciences / MEGVII Technology / Aberystwyth Universit
Compared to convolutional layers, fully-connected (FC) layers are better at modeling the long-range dependencies but worse at capturing the local patterns, hence usually less favored for image recognition. In this paper, we propose a methodology, Locality Injection, to incorporate local priors into an FC layer via merging the trained parameters of a parallel conv kernel into the FC kernel. Locality Injection can be viewed as a novel Structural Re-parameterization method since it equivalently converts the structures via transforming the parameters. Based on that, we propose a multi-layer-perceptron (MLP) block named RepMLP Block, which uses three FC layers to extract features, and a novel architecture named RepMLPNet. The hierarchical design distinguishes RepMLPNet from the other concurrently proposed vision MLPs. As it produces feature maps of different levels, it qualifies as a backbone model for downstream tasks like semantic segmentation. Our results reveal that 1) Locality Injection is a general methodology for MLP models; 2) RepMLPNet has favorable accuracy-efficiency trade-off compared to the other MLPs; 3) RepMLPNet is the first MLP that seamlessly transfer to Cityscapes semantic segmentation.

2022 Q1


Paper When Shift Operation Meets Vision Transformer: An Extremely Simple Alternative to Attention Mechanism
Date 26 Jan 2022
Code http://github.com/microsoft/SPACH
Authors Guangting Wang, Yucheng Zhao, Chuanxin Tang, Chong Luo, Wenjun Zeng
Org. University of Science and Technology of China / Microsoft Research Asia
Attention mechanism has been widely believed as the key to success of vision transformers (ViTs), since it provides a flexible and powerful way to model spatial relationships. However, is the attention mechanism truly an indispensable part of ViT? Can it be replaced by some other alternatives? To demystify the role of attention mechanism, we simplify it into an extremely simple case: ZERO FLOP and ZERO parameter. Concretely, we revisit the shift operation. It does not contain any parameter or arithmetic calculation. The only operation is to exchange a small portion of the channels between neighboring features. Based on this simple operation, we construct a new backbone network, namely ShiftViT, where the attention layers in ViT are substituted by shift operations. Surprisingly, ShiftViT works quite well in several mainstream tasks, e.g., classification, detection, and segmentation. The performance is on par with or even better than the strong baseline Swin Transformer. These results suggest that the attention mechanism might not be the vital factor that makes ViT successful. It can be even replaced by a zero-parameter operation. We should pay more attentions to the remaining parts of ViT in the future work.


Paper DynaMixer: A Vision MLP Architecture with Dynamic Mixing
Date 28 Jan 2022
Authors Ziyu Wang, Wenhao Jiang, Yiming Zhu, Li Yuan, Yibing Song, Wei Liu
Org. Tencent , Tsinghua University , Peking University
Recently, MLP-like vision models have achieved promising performances on mainstream visual recognition tasks. In contrast with vision transformers and CNNs, the success of MLP-like models shows that simple information fusion operations among tokens and channels can yield a good representation power for deep recognition models. However, existing MLP-like models fuse tokens through static fusion operations, lacking adaptability to the contents of the tokens to be mixed. Thus, customary information fusion procedures are not effective enough. To this end, this paper presents an efficient MLP-like network architecture, dubbed DynaMixer, resorting to dynamic information fusion. Critically, we propose a procedure, on which the DynaMixer model relies, to dynamically generate mixing matrices by leveraging the contents of all the tokens to be mixed. To reduce the time complexity and improve the robustness, a dimensionality reduction technique and a multi-segment fusion mechanism are adopted. Our proposed DynaMixer model (97M parameters) achieves 84.3% top-1 accuracy on the ImageNet-1K dataset without extra training data, performing favorably against the state-of-the-art vision MLP models. When the number of parameters is reduced to 26M, it still achieves 82.7% top-1 accuracy, surpassing the existing MLP-like models with a similar capacity.


Paper Mixing and Shifting: Exploiting Global and Local Dependencies in Vision MLPs
Date 14 Feb 2022
Code https://github.com/JegZheng/MS-MLP
Authors Huangjie Zheng, Pengcheng He, Weizhu Chen, Mingyuan Zhou
Org. The University of Texas / Microsoft Azure AI
Token-mixing multi-layer perceptron (MLP) models have shown competitive performance in computer vision tasks with a simple architecture and relatively small computational cost. Their success in maintaining computation efficiency is mainly attributed to avoiding the use of self-attention that is often computationally heavy, yet this is at the expense of not being able to mix tokens both globally and locally. In this paper, to exploit both global and local dependencies without self-attention, we present Mix-Shift-MLP (MS-MLP) which makes the size of the local receptive field used for mixing increase with respect to the amount of spatial shifting. In addition to conventional mixing and shifting techniques, MS-MLP mixes both neighboring and distant tokens from fine- to coarse-grained levels and then gathers them via a shifting operation. This directly contributes to the interactions between global and local tokens. Being simple to implement, MS-MLP achieves competitive performance in multiple vision benchmarks. For example, an MS-MLP with 85 million parameters achieves 83.8% top-1 classification accuracy on ImageNet-1K. Moreover, by combining MS-MLP with state-of-the-art Vision Transformers such as the Swin Transformer, we show MS-MLP achieves further improvements on three different model scales, e.g., by 0.5% on ImageNet-1K classification with Swin-B.


Paper ActiveMLP: An MLP-like Architecture with Active Token Mixer
Date 11 Mar 2022
Code https://github.com/microsoft/ActiveMLP
Authors Guoqiang Wei, Zhizheng Zhang, Cuiling Lan, Yan Lu, Zhibo Chen
Org. Microsoft Research Asia , University of Science and Technology of China
This paper presents ActiveMLP, a general MLP-like backbone for computer vision. The three existing dominant network families, i.e., CNNs, Transformers and MLPs, differ from each other mainly in the ways to fuse contextual information into a given token, leaving the design of more effective token-mixing mechanisms at the core of backbone architecture development. In ActiveMLP, we propose an innovative token-mixer, dubbed Active Token Mixer (ATM), to actively incorporate contextual information from other tokens in the global scope into the given one. This fundamental operator actively predicts where to capture useful contexts and learns how to fuse the captured contexts with the original information of the given token at channel levels. In this way, the spatial range of token-mixing is expanded and the way of token-mixing is reformed. With this design, ActiveMLP is endowed with the merits of global receptive fields and more flexible content-adaptive information fusion. Extensive experiments demonstrate that ActiveMLP is generally applicable and comprehensively surpasses different families of SOTA vision backbones by a clear margin on a broad range of vision tasks, including visual recognition and dense prediction tasks.

2022 Q2


Paper MDMLP: Image Classification from Scratch on Small Datasets with MLP
Date 28 May 2022
Code https://github.com/Amoza-Theodore/MDMLP
Authors Tian Lv, Chongyang Bai, Chaojie Wang
Org. Jiangsu University , Dartmouth College
The attention mechanism has become a go-to technique for natural language processing and computer vision tasks. Recently, the MLP-Mixer and other MLP-based architectures, based simply on multi-layer perceptrons (MLPs), are also powerful compared to CNNs and attention techniques and raises a new research direction. However, the high capability of the MLP-based networks severely relies on large volume of training data, and lacks of explanation ability compared to the Vision Transformer (ViT) or ConvNets. When trained on small datasets, they usually achieved inferior results than ConvNets. To resolve it, we present (i) multi-dimensional MLP (MDMLP), a conceptually simple and lightweight MLP-based architecture yet achieves SOTA when training from scratch on small-size datasets; (ii) multi-dimension MLP Attention Tool (MDAttnTool), a novel and efficient attention mechanism based on MLPs. Even without strong data augmentation, MDMLP achieves 90.90% accuracy on CIFAR10 with only 0.3M parameters, while the well-known MLP-Mixer achieves 85.45% with 17.1M parameters. In addition, the lightweight MDAttnTool highlights objects in images, indicating its explanation power.


Paper Parameterization of Cross-Token Relations with Relative Positional Encoding for Vision MLP
Date 15 Jul 2022
Code https://github.com/Zhicaiwww/PosMLP
Authors Zhicai Wang, Yanbin Hao, Xingyu Gao, Hao Zhang, Shuo Wang, Tingting Mu, Xiangnan He
Org. University of Science and Technology of China / Institute of Microelectronics, Chinese Academy of Sciences / Singapore Management University / University of Manchester
Vision multi-layer perceptrons (MLPs) have shown promising performance in computer vision tasks, and become the main competitor of CNNs and vision Transformers. They use token-mixing layers to capture cross-token interactions, as opposed to the multi-head self-attention mechanism used by Transformers. However, the heavily parameterized token-mixing layers naturally lack mechanisms to capture local information and multi-granular non-local relations, thus their discriminative power is restrained. To tackle this issue, we propose a new positional spacial gating unit (PoSGU). It exploits the attention formulations used in the classical relative positional encoding (RPE), to efficiently encode the cross-token relations for token mixing. It can successfully reduce the current quadratic parameter complexity O(N2) of vision MLPs to O(N) and O(1). We experiment with two RPE mechanisms, and further propose a group-wise extension to improve their expressive power with the accomplishment of multi-granular contexts. These then serve as the key building blocks of a new type of vision MLP, referred to as PosMLP. We evaluate the effectiveness of the proposed approach by conducting thorough experiments, demonstrating an improved or comparable performance with reduced parameter complexity. For instance, for a model trained on ImageNet1K, we achieve a performance improvement from 72.14% to 74.02% and a learnable parameter reduction from 19.4M to 18.2M.


Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up