Searching for MobileNetV3

We present the next generation of MobileNets based on a combination of complementary search techniques as well as a novel architecture design. MobileNetV3 is tuned to mobile phone CPUs through a combination of hardware-aware network architecture search (NAS) complemented by the NetAdapt algorithm and then subsequently improved through novel architecture advances... (read more)

PDF Abstract ICCV 2019 PDF ICCV 2019 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Semantic Segmentation Cityscapes test MobileNet V3-Large 1.0 Mean IoU (class) 72.6% # 43
Image Classification ImageNet MobileNet V3-Large 1.0 Top 1 Accuracy 75.2% # 162
Number of params 5.4M # 138

Methods used in the Paper


METHOD TYPE
Sigmoid Activation
Activation Functions
Dense Connections
Feedforward Networks
Squeeze-and-Excitation Block
Image Model Blocks
Global Average Pooling
Pooling Operations
ReLU
Activation Functions
Dropout
Regularization
ReLU6
Activation Functions
NetAdapt
Network Shrinking
Step Decay
Learning Rate Schedules
Random Horizontal Flip
Image Data Augmentation
Random Resized Crop
Image Data Augmentation
Weight Decay
Regularization
RMSProp
Stochastic Optimization
Hard Swish
Activation Functions
MobileNetV3
Convolutional Neural Networks