ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices
By Xiangyu Zhang, Xinyu Zhou, Mengxiao Lin. 2017
8,533 citations recorded by the source corpus.
We introduce an extremely computation-efficient CNN architecture named ShuffleNet, which is designed specially for mobile devices with very limited computing power (e.g., 10-150 MFLOPs). The new architecture utilizes two new operations, pointwise group convolution and channel shuffle, to greatly reduce computation cost while maintaining accuracy. Experiments on ImageNet classification and MS COCO object detection demonstrate the superior performance of ShuffleNet over other structures, e.g. lower top-1 error (absolute 7.8%) than recent MobileNet [12] on ImageNet classification task, under the computation budget of 40 MFLOPs. On an ARM-based mobile device, ShuffleNet achieves ~13× actual speedup over AlexNet while maintaining comparable accuracy.
Subfields: Computer Vision and Pattern Recognition, Efficiency & Compression
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Xiaogang Wang
- Ming-Hsuan Yang
- Trevor Darrell
- Yi Yang
- Alan L. Yuille
- Jitendra Malik
In Efficiency & Compression
Citations in the corpus
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks (2019, background): “As mobile phones become ubiquitous, it is also common to hand-craft efficient mobile-size ConvNets, such as SqueezeNets (Iandola et al., 2016; Gholami et al., 2018), MobileNets (Howard et al., 2017; Sandler et al., 2018), and ShuffleNets (Zhang et al., 2018; Ma et al., 2018).”
- Visual Transformers: Token-based Image Representation and Processing for Computer Vision (2020, background): “Early work in this direction includes [23, 31, 13, 17, 32, 16, 48, 27, 43].”
- Group Normalization (2018, background): “ShuffleNet (Zhang et al. 2018) proposes a channel shuffle operation that permutes the axes of grouped features.”
- Designing Network Design Spaces (2020, neutral): “SHUFFLENET [28] 0.”
- Squeeze-and-Excitation Networks (2017, methodology): “0 MobileNet-224” in [13] and ShuffleNet refers to “ShuffleNet 1× (g = 3)” in [52].”
- Learning Transferable Architectures for Scalable Image Recognition (2017, methodology): “…address this question, we plug in the family of Model resolution mAP (mini-val) mAP (test-dev) MobileNet-224 [14] 600 × 600 19.8% - ShuffleNet (2x) [43] 600 × 600 24.5% † - NASNet-A (4 @ 1056) 600 × 600 29.6% - ResNet-101-FPN [24] 800 (short side) - 36.2% Inception-ResNet-v2 (G-RMI) [18] 600 × 600…”
- Green AI (2019, background): “Most of these approaches target efficient model inference [31, 49, 12],15 and thus only minimize the cost of processing a single example, while ignoring the other two red practices discussed in Section 2.”
- Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation (2018, methodology): “Motivated by the recent success of depthwise separable convolution [67, 71, 12, 31, 80], we also explore this operation and show improvement in terms of both speed and accuracy by adapting the Xception model [12], similar to”
- ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design (2018, methodology): “1 In this paper, the definition of FLOPs follows [35], i.”
- DARTS: Differentiable Architecture Search (2018, neutral): “2 569 – manual ShuffleNet 2× (g = 3) (Zhang et al., 2017) 26.”
- MnasNet: Platform-Aware Neural Architecture Search for Mobile (2018, background): “…and reducing filter sizes; MobileNet [11] extensively employs depthwise separable convolution to minimize computation density; ShuffleNets [33, 24] utilize low-cost group convolution and channel shuffle; Con-densenet [14] learns to connect group convolutions across layers; Recently,…”
- Gather-Excite: Exploiting Feature Context in Convolutional Neural Networks (2018, methodology): “To answer this question, we conduct a series of experiments on the ShuffleNet architecture [50], an efficient model that achieves a good tradeoff between accuracy and latency.”
- MiLeNAS: Efficient Neural Architecture Search via Mixed-Level Reformulation (2020, neutral): “2 569 manual ShuffleNet [31] 26.”
- Neural Architecture Optimization (2018, neutral): “2 ∼ 5 524 ShuffleNet 2× (v2) [49] 26.”
- Group Knowledge Transfer: Collaborative Training of Large CNNs on the Edge (2020, methodology): “Our work also relates to efficient deep learning on edge devices, such as model compression [43, 44, 45], manually designed architectures (MobileNets [46], ShuffeNets [47], SqueezeNets [48]), or even efficient neural architecture search (EfficientNets [4], FBNet [49]).”