资源算法 mobilenetv3-tensorflow-printer

mobilenetv3-tensorflow-printer

2020-02-28 | |  94 |   0 |   0

MobileNetV3 TensorFlow

Unofficial implementation of MobileNetV3 architecture described in paper Searching for MobileNetV3. This repository follows mobilenetv3-tensorflow.

Requirements

  • Python 3.6+

  • TensorFlow 1.13+

pip install -r requirements.txt

Build model

MobileNetV3 Small

from mobilenetv3_factory import build_mobilenetv3
model = build_mobilenetv3(    "small",    input_shape=(224, 224, 3),    num_classes=3,    width_multiplier=1.0,
)

MobileNetV3 Large

from mobilenetv3_factory import build_mobilenetv3
model = build_mobilenetv3(    "large",    input_shape=(224, 224, 3),    num_classes=3,    width_multiplier=1.0,
)

Train

Printer dataset

python train.py 
    --model_type small 
    --width_multiplier 1.0 
    --height 224 
    --width 224 
    --dataset printer 
    --lr 0.001 
    --optimizer sgd 
    --train_batch_size 10 
    --valid_batch_size 10 
    --num_epoch 30 
    --logdir logdir

Test

printer dataset

python test.py 
    --model_path mobilenetv3_small_printer_30.h5 
    --model_type small 
    --image_path test/1.jpg 
    --featuremap False 
    --heatmap False

TensorBoard

Graph, training and evaluaion metrics are saved to TensorBoard event file uder directory specified with --logdir` argument during training. You can launch TensorBoard using following command.

tensorboard --logdir logdir


上一篇:WindowsML-CoreML-MobileNet-demo

下一篇:maskrcnn_benchmark_mobilenetv2

用户评价
全部评价

热门资源

  • TensorFlow-Course

    This repository aims to provide simple and read...

  • seetafaceJNI

    项目介绍 基于中科院seetaface2进行封装的JAVA...

  • mxnet_VanillaCNN

    This is a mxnet implementation of the Vanilla C...

  • DuReader_QANet_BiDAF

    Machine Reading Comprehension on DuReader Usin...

  • Klukshu-Sockeye-...

    KLUKSHU SOCKEYE PROJECTS 2016 This repositor...