R-MnasNet: Reduced MnasNet for Computer Vision

Shah, Prasham; El-Sharkawy, Mohamed

doi:10.1109/iemtronics51293.2020.9216434

Cited by 3 publications

(1 citation statement)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…MNasNet is a mobile neural architecture search method that explicitly incorporates real-world inference latency into the main objective so that the search can identify an architecture that achieves an optimal accuracy-latency tradeoff [3]. Shah and El-Sharkawy introduce two variations of MNasNet named A-MNasnet and R-MNasnet in [4] and [5] respectively. The former introduced a new architecture based on MNasNet that makes the model more efficient in terms of accuracy, and the latter introduces a modified architecture based on MNasNet that makes it more compact with a fair trade-off between model size and accuracy.…”

Section: Related Workmentioning

confidence: 99%

Efficient Incorporation of Multiple Latency Targets in the Once-For-All Network

Kumar¹,

Andrew²

2020

Preprint

View full text Add to dashboard Cite

Neural Architecture Search has proven an effective method of automating architecture engineering. Recent work in the field has been to look for architectures subject to multiple objectives such as accuracy and latency to efficiently deploy them on different target hardware. Once-for-All (OFA) is one such method that decouples training and search and is able to find high-performance networks for different latency constraints. However, the search phase is inefficient at incorporating multiple latency targets. In this paper, we introduce two strategies (Top-down and Bottom-up) that use warm starting and randomized network pruning for the efficient incorporation of multiple latency targets in the OFA network. We evaluate these strategies against the current OFA implementation and demonstrate that our strategies offer significant running time performance gains while not sacrificing the accuracy of the subnetworks that were found for each latency target. We further demonstrate that these performance gains are generalized to every design space used by the OFA network. The code is available here.Preprint. Under review.

show abstract

Section: Related Workmentioning

confidence: 99%

Efficient Incorporation of Multiple Latency Targets in the Once-For-All Network

Kumar¹,

Andrew²

2020

Preprint

View full text Add to dashboard Cite

show abstract

Image Classification with A-MnasNet and R-MnasNet on NXP Bluebox 2.0

Shah

El-Sharkawy

2021

Lecture Notes in Networks and Systems

View full text Add to dashboard Cite

Multi-Attention Ghost Residual Fusion Network for Image Classification

Jia

Guo

et al. 2021

IEEE Access

View full text Add to dashboard Cite

In order to achieve high-efficiency and high-precision multi-image classification tasks, a multi-attention ghost residual fusion network (MAGR) is proposed. MAGR is formed by cascading basic feature extraction network (BFE), ghost residual mapping network (GRM) and image classification network (IC). The BFE uses spatial and channel attention mechanisms to help the MAGR extract low-level features of the input image in a targeted manner. The GRM is formed by cascading 4 multi-branch group convolutional ghost residual blocks (MGR-Blocks). Each MGR-Block is cascaded by a dimension reducer and several ghost residual sub-networks (GRSs). The GRS integrates ghost convolution and residual connection, and the use of ghost convolution can significantly reduce parameters and achieve high-efficient classification. The GRS is a parallel convolution structure with 32 branches, which ensures that GRM has enough width to extract advanced features and extract as much feature information as possible, so as to obtain high-precision classification. The IC completes the aggregation of high-dimensional channel feature information, and then achieves a significant improvement in the classification accuracy of MAGR, by fusing the effective channel attention mechanism, global average pooling and SoftMax layer. Simulation experiment shows that MAGR has excellent classification capability while achieving high efficiency and lightweight. Compare with VGG16, the parameters of MAGR on CIFAR-10 is reduced by 94.8% while the classification accuracy is increased by 1.18%. Compare with MobileNetV2, the parameters of MAGR on CIFAR-100 is reduced by 33.9% while the classification accuracy is increased by 15.6%.

show abstract

R-MnasNet: Reduced MnasNet for Computer Vision

Cited by 3 publications

References 13 publications

Efficient Incorporation of Multiple Latency Targets in the Once-For-All Network

Efficient Incorporation of Multiple Latency Targets in the Once-For-All Network

Image Classification with A-MnasNet and R-MnasNet on NXP Bluebox 2.0

Multi-Attention Ghost Residual Fusion Network for Image Classification

Contact Info

Product

Resources

About