Papers › Mobile-Former: Bridging MobileNet and Transformer

Mobile-Former: Bridging MobileNet and Transformer

12 Aug 2021CVPR 2022 1arXiv:2108.05895archive 2025-07-28

Yinpeng Chen, Xiyang Dai, Dongdong Chen, Mengchen Liu, Xiaoyi Dong, Lu Yuan, Zicheng Liu

We present Mobile-Former, a parallel design of MobileNet and transformer with a two-way bridge in between. This structure leverages the advantages of MobileNet at local processing and transformer at global interaction. And the bridge enables bidirectional fusion of local and global features. Different from recent works on vision transformer, the transformer in Mobile-Former contains very few tokens (e.g. 6 or fewer tokens) that are randomly initialized to learn global priors, resulting in low computational cost. Combining with the proposed light-weight cross attention to model the bridge, Mobile-Former is not only computationally efficient, but also has more representation power. It outperforms MobileNetV3 at low FLOP regime from 25M to 500M FLOPs on ImageNet classification. For instance, Mobile-Former achieves 77.9\% top-1 accuracy at 294M FLOPs, gaining 1.3\% over MobileNetV3 but saving 17\% of computations. When transferring to object detection, Mobile-Former outperforms MobileNetV3 by 8.6 AP in RetinaNet framework. Furthermore, we build an efficient end-to-end detector by replacing backbone, encoder and decoder in DETR with Mobile-Former, which outperforms DETR by 1.1 AP but saves 52\% of computational cost and 36\% of parameters.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

aaboys/mobileformer mentioned on GitHubpytorch report
federicopozzi33/MobileOne-PyTorch mentioned on GitHubpytorchApache-2.0 report
BR-IDL/PaddleViT paddleApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DecoderObject Detectionobject-detection

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

Methods

1x1 ConvolutionAbsolute Position EncodingsAdamAttentionAverage PoolingBPEBatch NormalizationConvolutionDense ConnectionsDepthwise ConvolutionDepthwise Separable ConvolutionDetrDropoutFPNFeedforward NetworkFocal LossGlobal Average PoolingHard SwishInverted Residual BlockLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPointwise ConvolutionPosition-Wise Feed-Forward LayerReLUReLU6Residual ConnectionRetinaNetSigmoid ActivationSoftmaxSqueeze-and-Excitation BlockTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections