Sandbox for training large-scale image classification networks for embedded systems
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
chainer_ Update PIP-packages Dec 15, 2018
common Work on InceptionV3, 3 Oct 29, 2018
datasets Working on OI4C1K script, 5 Nov 16, 2018
gluon KHPA: Move to BCE/F1 loss/metric Dec 15, 2018
keras_ Update PIP-packages Dec 15, 2018
pytorch Update PIP-packages Dec 15, 2018
tensorflow_ Update PIP-packages Dec 15, 2018
tests Start working on Xception Nov 1, 2018
.gitignore Update gitignore Jul 9, 2018
.travis.yml Fix some PEP8 warnings Dec 7, 2018
LICENSE Initial commit Jul 9, 2018
README.md Update info about MobileNet x0.75 models Dec 15, 2018
__init__.py Working on pretrained model loading Aug 9, 2018
convert_models.py Fix pep8 warnings Dec 12, 2018
eval_ch.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
eval_gl.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
eval_gl_khpa.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
eval_ke.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
eval_pt.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
eval_tf.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
requirements.txt Working on KHPA DS, 6 Dec 13, 2018
train_ch.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
train_gl.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
train_gl_khpa.py KHPA: Fix selecting the best model for F1 metric Dec 15, 2018
train_ke.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
train_pt.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018
train_tf.py 1) change comments, 2) KHPA: Some experiments with various loss/metrics Dec 15, 2018

README.md

Large-scale image classification networks for embedded systems

Build Status GitHub License Python Version

This repository contains several classification models on MXNet/Gluon, PyTorch, Chainer, Keras, and TensorFlow, with scripts for training/validating/converting models. All models are designed for using with ImageNet-1k dataset.

List of models

Installation

For Gluon way

To use only Gluon models in your project, simply install the gluoncv2 package with mxnet:

pip install gluoncv2 mxnet>=1.2.1

To enable different hardware supports such as GPUs, check out MXNet variants. For example, you can install with CUDA-9.2 supported MXNet:

pip install gluoncv2 mxnet-cu92>=1.2.1

For PyTorch way

To use only PyTorch models in your project, simply install the pytorchcv package with torch (>=0.4.1 is recommended):

pip install pytorchcv torch>=0.4.0

To enable/disable different hardware supports such as GPUs, check out PyTorch installation instructions.

For Chainer way

To use only Chainer models in your project, simply install the chainercv2 package:

pip install chainercv2

For Keras way

To use only Keras models in your project, simply install the kerascv package with mxnet:

pip install kerascv mxnet>=1.2.1

To enable different hardware supports such as GPUs, check out MXNet variants. For example, you can install with CUDA-9.2 supported MXNet:

pip install kerascv mxnet-cu92>=1.2.1

After installation change the value of the field image_data_format to channels_first in the file ~/.keras/keras.json.

For TensorFlow way

To use only TensorFlow models in your project, simply install the tensorflowcv package with tensorflow-gpu:

pip install tensorflowcv tensorflow-gpu>=1.11.0

To enable/disable different hardware supports, check out TensorFlow installation instructions.

Note that the models use NCHW data format. The current version of TensorFlow cannot work with them on CPU.

For research

To use the repository for training/validation/converting models:

git clone git@github.com:osmr/imgclsmob.git
pip install -r requirements.txt

Usage

For Gluon way

Example of using the pretrained ResNet-18 model on Gluon:

from gluoncv2.model_provider import get_model as glcv2_get_model
import mxnet as mx

net = glcv2_get_model("resnet18", pretrained=True)
x = mx.nd.zeros((1, 3, 224, 224), ctx=mx.cpu())
y = net(x)

For PyTorch way

Example of using the pretrained ResNet-18 model on PyTorch:

from pytorchcv.model_provider import get_model as ptcv_get_model
import torch
from torch.autograd import Variable

net = ptcv_get_model("resnet18", pretrained=True)
x = Variable(torch.randn(1, 3, 224, 224))
y = net(x)

For Chainer way

Example of using the pretrained ResNet-18 model on Chainer:

from chainercv2.model_provider import get_model as chcv2_get_model
import numpy as np

net = chcv2_get_model("resnet18", pretrained=True)
x = np.zeros((1, 3, 224, 224), np.float32)
y = net(x)

For Keras way

Example of using the pretrained ResNet-18 model on Keras:

from kerascv.model_provider import get_model as kecv_get_model
import numpy as np

net = kecv_get_model("resnet18", pretrained=True)
x = np.zeros((1, 3, 224, 224), np.float32)
y = net.predict(x)

For TensorFlow way

Example of using the pretrained ResNet-18 model on TensorFlow:

from tensorflowcv.model_provider import get_model as tfcv_get_model
from tensorflowcv.model_provider import init_variables_from_state_dict as tfcv_init_variables_from_state_dict
import tensorflow as tf
import numpy as np

net = tfcv_get_model("resnet18", pretrained=True)
x = tf.placeholder(dtype=tf.float32, shape=(None, 3, 224, 224), name='xx')
y_net = net(x)

with tf.Session() as sess:
    tfcv_init_variables_from_state_dict(sess=sess, state_dict=net.state_dict)
    x_value = np.zeros((1, 3, 224, 224), np.float32)
    y = sess.run(y_net, feed_dict={x: x_value})

Pretrained models

Some remarks:

  • All pretrained models can be downloaded automatically during use (use the parameter pretrained).
  • Top1/Top5 are the standard 1-crop Top-1/Top-5 errors (in percents) on the validation subset of the ImageNet1k dataset.
  • ResNet/PreResNet with b-suffix is a version of the networks with the stride in the second convolution of the bottleneck block. Respectively a network without b-suffix has the stride in the first convolution.
  • ResNet/PreResNet models do not use biases in convolutions at all.
  • CondenseNet models are only so-called converted versions.
  • ShuffleNetV2/ShuffleNetV2b/ShuffleNetV2c are different implementations of the same architecture.
  • All models require ordinary normalization.

For Gluon

Model Top1 Top5 Params FLOPs Remarks
AlexNet 44.12 21.26 61,100,840 715.49M From dmlc/gluon-cv (log)
VGG-11 31.91 11.76 132,863,336 7,622.65M From dmlc/gluon-cv (log)
VGG-13 31.06 11.12 133,047,848 11,326.85M From dmlc/gluon-cv (log)
VGG-16 26.78 8.69 138,357,544 15,489.95M From dmlc/gluon-cv (log)
VGG-19 25.88 8.23 143,667,240 19,653.05M From dmlc/gluon-cv (log)
BN-VGG-11b 30.34 10.57 132,868,840 7,622.65M From dmlc/gluon-cv (log)
BN-VGG-13b 29.48 10.16 133,053,736 11,326.85M From dmlc/gluon-cv (log)
BN-VGG-16b 26.89 8.65 138,365,992 15,489.95M From dmlc/gluon-cv (log)
BN-VGG-19b 25.66 8.15 143,678,248 19,653.05M From dmlc/gluon-cv (log)
ResNet-10 37.09 15.55 5,418,792 892.62M Training (log)
ResNet-12 35.86 14.46 5,492,776 1,124.23M Training (log)
ResNet-14 32.85 12.41 5,788,200 1,355.64M Training (log)
ResNet-16 30.68 11.10 6,968,872 1,586.95M Training (log)
ResNet-18 x0.25 49.16 24.45 831,096 136.64M Training (log)
ResNet-18 x0.5 36.54 14.96 3,055,880 485.22M Training (log)
ResNet-18 x0.75 33.25 12.54 6,675,352 1,045.75M Training (log)
ResNet-18 29.13 9.94 11,689,512 1,818.21M Training (log)
ResNet-34 25.34 7.92 21,797,672 3,669.16M From dmlc/gluon-cv (log)
ResNet-50 23.50 6.87 25,557,032 3,868.96M From dmlc/gluon-cv (log)
ResNet-50b 22.92 6.44 25,557,032 4,100.70M From dmlc/gluon-cv (log)
ResNet-101 21.66 5.99 44,549,160 7,586.30M From dmlc/gluon-cv (log)
ResNet-101b 21.18 5.60 44,549,160 7,818.04M From dmlc/gluon-cv (log)
ResNet-152 21.01 5.61 60,192,808 11,304.85M From dmlc/gluon-cv (log)
ResNet-152b 20.54 5.37 60,192,808 11,536.58M From dmlc/gluon-cv (log)
PreResNet-18 28.72 9.88 11,687,848 1,818.41M Training (log)
PreResNet-34 25.88 8.11 21,796,008 3,669.36M From dmlc/gluon-cv (log)
PreResNet-50 23.39 6.68 25,549,480 3,869.16M From dmlc/gluon-cv (log)
PreResNet-50b 23.16 6.64 25,549,480 4,100.90M From dmlc/gluon-cv (log)
PreResNet-101 21.45 5.75 44,541,608 7,586.50M From dmlc/gluon-cv (log)
PreResNet-101b 21.73 5.88 44,541,608 7,818.24M From dmlc/gluon-cv (log)
PreResNet-152 20.70 5.32 60,185,256 11,305.05M From dmlc/gluon-cv (log)
PreResNet-152b 21.00 5.75 60,185,256 11,536.78M From dmlc/gluon-cv (log)
PreResNet-200b 21.10 5.64 64,666,280 15,040.27M From tornadomeet/ResNet (log)
ResNeXt-101 (32x4d) 21.32 5.79 44,177,704 7,991.62M From Cadene/pretrained...pytorch (log)
ResNeXt-101 (64x4d) 20.60 5.41 83,455,272 15,491.88M From Cadene/pretrained...pytorch (log)
SE-ResNet-50 22.51 6.44 28,088,024 3,877.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-101 21.92 5.89 49,326,872 7,600.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-152 21.48 5.77 66,821,848 11,324.62M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-50 (32x4d) 21.06 5.58 27,559,896 4,253.33M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-101 (32x4d) 19.99 5.00 48,955,416 8,005.33M From Cadene/pretrained...pytorch (log)
SENet-154 18.84 4.65 115,088,984 20,742.40M From Cadene/pretrained...pytorch (log)
IBN-ResNet-50 23.56 6.68 25,557,032 4,100.70M From XingangPan/IBN-Net (log)
IBN-ResNet-101 21.89 5.87 44,549,160 7,818.04M From XingangPan/IBN-Net (log)
IBN(b)-ResNet-50 23.91 6.97 25,558,568 4,100.70M From XingangPan/IBN-Net (log)
IBN-ResNeXt-101 (32x4d) 21.43 5.62 44,177,704 7,991.62M From XingangPan/IBN-Net (log)
IBN-DenseNet-121 24.98 7.47 7,978,856 2,852.39M From XingangPan/IBN-Net (log)
IBN-DenseNet-169 23.78 6.82 14,149,480 3,381.48M From XingangPan/IBN-Net (log)
AirNet50-1x64d (r=2) 22.48 6.21 27,425,864 4,757.77M From soeaver/AirNet-PyTorch (log)
AirNet50-1x64d (r=16) 22.91 6.46 25,714,952 4,385.54M From soeaver/AirNet-PyTorch (log)
BAM-ResNet-50 23.68 6.96 25,915,099 4,211.08M From Jongchan/attention-module (log)
CBAM-ResNet-50 23.02 6.38 28,089,624 4,118.19M From Jongchan/attention-module (log)
PyramidNet-101 (a=360) 22.72 6.52 42,455,070 8,706.81M From dyhan0920/Pyramid...PyTorch (log)
DiracNetV2-18 30.61 11.17 11,511,784 1,798.43M From szagoruyko/diracnets (log)
DiracNetV2-34 27.93 9.46 21,616,232 3,649.37M From szagoruyko/diracnets (log)
DenseNet-121 25.11 7.80 7,978,856 2,852.39M From dmlc/gluon-cv (log)
DenseNet-161 22.40 6.18 28,681,000 7,761.25M From dmlc/gluon-cv (log)
DenseNet-169 23.89 6.89 14,149,480 3,381.48M From dmlc/gluon-cv (log)
DenseNet-201 22.71 6.36 20,013,928 4,318.75M From dmlc/gluon-cv (log)
CondenseNet-74 (C=G=4) 26.82 8.64 4,773,944 533.64M From ShichenLiu/CondenseNet (log)
CondenseNet-74 (C=G=8) 29.76 10.49 2,935,416 278.55M From ShichenLiu/CondenseNet (log)
WRN-50-2 22.15 6.12 68,849,128 11,412.82M From szagoruyko/functional-zoo (log)
DRN-C-26 25.68 7.89 21,126,584 20,838.70M From fyu/drn (log)
DRN-C-42 23.80 6.92 31,234,744 31,236.97M From fyu/drn (log)
DRN-C-58 22.35 6.27 40,542,008 36,862.32M From fyu/drn (log)
DRN-D-22 26.67 8.52 16,393,752 16,626.00M From fyu/drn (log)
DRN-D-38 24.51 7.36 26,501,912 27,024.27M From fyu/drn (log)
DRN-D-54 22.05 6.27 35,809,176 32,649.62M From fyu/drn (log)
DRN-D-105 21.31 5.81 54,801,304 48,682.11M From fyu/drn (log)
DPN-68 23.57 7.00 12,611,602 2,338.71M From Cadene/pretrained...pytorch (log)
DPN-98 20.23 5.28 61,570,728 11,702.80M From Cadene/pretrained...pytorch (log)
DPN-131 20.03 5.22 79,254,504 16,056.22M From Cadene/pretrained...pytorch (log)
DarkNet Tiny 40.31 17.46 1,042,104 496.34M Training (log)
DarkNet Ref 38.00 16.68 7,319,416 365.55M Training (log)
SqueezeNet v1.0 38.73 17.34 1,248,424 828.30M Training (log)
SqueezeNet v1.1 39.09 17.39 1,235,496 354.88M Training (log)
SqueezeResNet v1.1 39.83 17.84 1,235,496 354.88M Training (log)
ShuffleNetV2 x0.5 40.61 18.30 1,366,792 42.34M Training (log)
ShuffleNetV2b x0.5 40.98 18.56 1,366,792 42.37M Training (log)
ShuffleNetV2c x0.5 39.87 18.11 1,366,792 42.37M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.0 33.76 13.22 2,278,604 147.92M Training (log)
ShuffleNetV2c x1.0 30.74 11.38 2,279,760 148.85M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.5 32.38 12.37 4,406,098 318.61M Training (log)
ShuffleNetV2 x2.0 32.04 12.10 7,601,686 593.66M Training (log)
108-MENet-8x1 (g=3) 43.62 20.30 654,516 40.64M Training (log)
128-MENet-8x1 (g=4) 42.10 19.13 750,796 43.58M Training (log)
228-MENet-12x1 (g=3) 35.03 13.99 1,806,568 148.93M From clavichord93/MENet (log)
256-MENet-12x1 (g=4) 34.49 13.90 1,888,240 146.11M From clavichord93/MENet (log)
348-MENet-12x1 (g=3) 31.17 11.41 3,368,128 306.31M From clavichord93/MENet (log)
352-MENet-12x1 (g=8) 34.70 13.75 2,272,872 151.03M From clavichord93/MENet (log)
456-MENet-24x1 (g=3) 29.57 10.43 5,304,784 560.72M From clavichord93/MENet (log)
MobileNet x0.25 45.78 22.18 470,072 42.30M Training (log)
MobileNet x0.5 36.12 14.81 1,331,592 152.04M Training (log)
MobileNet x0.75 29.85 10.51 2,585,560 329.22M Training (log)
MobileNet x1.0 29.25 10.03 4,231,976 573.83M From dmlc/gluon-cv (log)
FD-MobileNet x0.25 56.19 31.38 383,160 12.44M Training (log)
FD-MobileNet x0.5 42.62 19.69 993,928 40.93M Training (log)
FD-MobileNet x1.0 34.42 13.73 2,901,288 146.08M Training (log)
MobileNetV2 x0.25 48.89 25.24 1,516,392 32.22M From dmlc/gluon-cv (log)
MobileNetV2 x0.5 35.51 14.64 1,964,736 95.62M From dmlc/gluon-cv (log)
MobileNetV2 x0.75 30.82 11.26 2,627,592 191.61M From dmlc/gluon-cv (log)
MobileNetV2 x1.0 28.51 9.90 3,504,960 320.19M From dmlc/gluon-cv (log)
IGCV3 28.22 9.54 3,491,688 323.81M From homles11/IGCV3 (log)
MnasNet 31.32 11.44 4,308,816 310.75M From zeusees/Mnasnet...Model (log)
DARTS 27.23 8.97 4,718,752 537.64M From quark0/darts (log)
Xception 20.99 5.56 22,855,952 8,385.86M From Cadene/pretrained...pytorch (log)
InceptionV3 21.22 5.59 23,834,568 5,746.72M From dmlc/gluon-cv (log)
InceptionV4 20.60 5.25 42,679,816 12,314.17M From Cadene/pretrained...pytorch (log)
InceptionResNetV2 19.96 4.94 55,843,464 13,189.58M From Cadene/pretrained...pytorch (log)
PolyNet 19.09 4.53 95,366,600 34,768.84M From Cadene/pretrained...pytorch (log)
NASNet-A 4@1056 25.37 7.95 5,289,978 587.29M From Cadene/pretrained...pytorch (log)
NASNet-A 6@4032 18.17 4.24 88,753,150 24,021.18M From Cadene/pretrained...pytorch (log)
PNASNet-5-Large 17.90 4.28 86,057,668 25,169.47M From Cadene/pretrained...pytorch (log)

For PyTorch

Model Top1 Top5 Params FLOPs Remarks
AlexNet 43.48 20.93 61,100,840 715.49M From dmlc/gluon-cv (log)
VGG-11 30.98 11.37 132,863,336 7,622.65M From dmlc/gluon-cv (log)
VGG-13 30.07 10.75 133,047,848 11,326.85M From dmlc/gluon-cv (log)
VGG-16 27.15 8.92 138,357,544 15,489.95M From dmlc/gluon-cv (log)
VGG-19 26.19 8.39 143,667,240 19,653.05M From dmlc/gluon-cv (log)
BN-VGG-11b 29.63 10.19 132,868,840 7,622.65M From dmlc/gluon-cv (log)
BN-VGG-13b 28.41 9.63 133,053,736 11,326.85M From dmlc/gluon-cv (log)
BN-VGG-16b 27.19 8.74 138,365,992 15,489.95M From dmlc/gluon-cv (log)
BN-VGG-19b 26.06 8.40 143,678,248 19,653.05M From dmlc/gluon-cv (log)
ResNet-10 37.46 15.85 5,418,792 892.62M Converted from GL model (log)
ResNet-12 36.18 14.80 5,492,776 1,124.23M Converted from GL model (log)
ResNet-14 33.17 12.71 5,788,200 1,355.64M Converted from GL model (log)
ResNet-16 30.90 11.38 6,968,872 1,586.95M Converted from GL model (log)
ResNet-18 x0.25 49.50 24.83 831,096 136.64M Converted from GL model (log)
ResNet-18 x0.5 37.04 15.38 3,055,880 485.22M Converted from GL model (log)
ResNet-18 x0.75 33.61 12.85 6,675,352 1,045.75M Converted from GL model (log)
ResNet-18 29.52 10.21 11,689,512 1,818.21M Converted from GL model (log)
ResNet-34 25.66 8.18 21,797,672 3,669.16M From dmlc/gluon-cv (log)
ResNet-50 23.79 7.05 25,557,032 3,868.96M From dmlc/gluon-cv (log)
ResNet-50b 23.05 6.65 25,557,032 4,100.70M From dmlc/gluon-cv (log)
ResNet-101 21.90 6.22 44,549,160 7,586.30M From dmlc/gluon-cv (log)
ResNet-101b 21.45 5.81 44,549,160 7,818.04M From dmlc/gluon-cv (log)
ResNet-152 21.26 5.82 60,192,808 11,304.85M From dmlc/gluon-cv (log)
ResNet-152b 20.74 5.50 60,192,808 11,536.58M From dmlc/gluon-cv (log)
PreResNet-18 29.09 10.18 11,687,848 1,818.41M Converted from GL model (log)
PreResNet-34 26.23 8.41 21,796,008 3,669.36M From dmlc/gluon-cv (log)
PreResNet-50 23.70 6.85 25,549,480 3,869.16M From dmlc/gluon-cv (log)
PreResNet-50b 23.33 6.87 25,549,480 4,100.90M From dmlc/gluon-cv (log)
PreResNet-101 21.74 5.91 44,541,608 7,586.50M From dmlc/gluon-cv (log)
PreResNet-101b 21.95 6.03 44,541,608 7,818.24M From dmlc/gluon-cv (log)
PreResNet-152 20.94 5.55 60,185,256 11,305.05M From dmlc/gluon-cv (log)
PreResNet-152b 21.34 5.91 60,185,256 11,536.78M From dmlc/gluon-cv (log)
PreResNet-200b 21.33 5.88 64,666,280 15,040.27M From tornadomeet/ResNet (log)
ResNeXt-101 (32x4d) 21.81 6.11 44,177,704 7,991.62M From Cadene/pretrained...pytorch (log)
ResNeXt-101 (64x4d) 21.04 5.75 83,455,272 15,491.88M From Cadene/pretrained...pytorch (log)
SE-ResNet-50 22.47 6.40 28,088,024 3,877.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-101 21.88 5.89 49,326,872 7,600.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-152 21.48 5.76 66,821,848 11,324.62M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-50 (32x4d) 21.00 5.54 27,559,896 4,253.33M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-101 (32x4d) 19.96 5.05 48,955,416 8,005.33M From Cadene/pretrained...pytorch (log)
SENet-154 18.62 4.61 115,088,984 20,742.40M From Cadene/pretrained...pytorch (log)
IBN-ResNet-50 22.76 6.41 25,557,032 4,100.70M From XingangPan/IBN-Net (log)
IBN-ResNet-101 21.29 5.61 44,549,160 7,818.04M From XingangPan/IBN-Net (log)
IBN(b)-ResNet-50 23.64 6.86 25,558,568 4,100.70M From XingangPan/IBN-Net (log)
IBN-ResNeXt-101 (32x4d) 20.88 5.42 44,177,704 7,991.62M From XingangPan/IBN-Net (log)
IBN-DenseNet-121 24.47 7.25 7,978,856 2,852.39M From XingangPan/IBN-Net (log)
IBN-DenseNet-169 23.25 6.51 14,149,480 3,381.48M From XingangPan/IBN-Net (log)
AirNet50-1x64d (r=2) 21.84 5.90 27,425,864 4,757.77M From soeaver/AirNet-PyTorch (log)
AirNet50-1x64d (r=16) 22.11 6.19 25,714,952 4,385.54M From soeaver/AirNet-PyTorch (log)
AirNeXt50-32x4d (r=2) 20.87 5.51 27,604,296 5,321.18M From soeaver/AirNet-PyTorch (log)
BAM-ResNet-50 23.14 6.58 25,915,099 4,211.08M From Jongchan/attention-module (log)
CBAM-ResNet-50 22.38 6.05 28,089,624 4,118.19M From Jongchan/attention-module (log)
PyramidNet-101 (a=360) 21.98 6.20 42,455,070 8,706.81M From dyhan0920/Pyramid...PyTorch (log)
DiracNetV2-18 31.47 11.70 11,511,784 1,798.43M From szagoruyko/diracnets (log)
DiracNetV2-34 28.75 9.93 21,616,232 3,649.37M From szagoruyko/diracnets (log)
DenseNet-121 25.57 8.03 7,978,856 2,852.39M From dmlc/gluon-cv (log)
DenseNet-161 22.86 6.44 28,681,000 7,761.25M From dmlc/gluon-cv (log)
DenseNet-169 24.40 7.19 14,149,480 3,381.48M From dmlc/gluon-cv (log)
DenseNet-201 23.10 6.63 20,013,928 4,318.75M From dmlc/gluon-cv (log)
CondenseNet-74 (C=G=4) 26.25 8.28 4,773,944 533.64M From ShichenLiu/CondenseNet (log)
CondenseNet-74 (C=G=8) 28.93 10.06 2,935,416 278.55M From ShichenLiu/CondenseNet (log)
WRN-50-2 22.53 6.41 68,849,128 11,412.82M From szagoruyko/functional-zoo (log)
DRN-C-26 24.86 7.55 21,126,584 20,838.70M From fyu/drn (log)
DRN-C-42 22.94 6.57 31,234,744 31,236.97M From fyu/drn (log)
DRN-C-58 21.73 6.01 40,542,008 36,862.32M From fyu/drn (log)
DRN-D-22 25.80 8.23 16,393,752 16,626.00M From fyu/drn (log)
DRN-D-38 23.79 6.95 26,501,912 27,024.27M From fyu/drn (log)
DRN-D-54 21.22 5.86 35,809,176 32,649.62M From fyu/drn (log)
DRN-D-105 20.62 5.48 54,801,304 48,682.11M From fyu/drn (log)
DPN-68 24.17 7.27 12,611,602 2,338.71M From Cadene/pretrained...pytorch (log)
DPN-98 20.81 5.53 61,570,728 11,702.80M From Cadene/pretrained...pytorch (log)
DPN-131 20.54 5.48 79,254,504 16,056.22M From Cadene/pretrained...pytorch (log)
DarkNet Tiny 40.74 17.84 1,042,104 496.34M Converted from GL model (log)
DarkNet Ref 38.58 17.18 7,319,416 365.55M Converted from GL model (log)
SqueezeNet v1.0 39.29 17.66 1,248,424 828.30M Converted from GL model (log)
SqueezeNet v1.1 39.31 17.72 1,235,496 354.88M Converted from GL model (log)
SqueezeResNet v1.1 40.09 18.21 1,235,496 354.88M Converted from GL model (log)
ShuffleNetV2 x0.5 40.99 18.65 1,366,792 42.34M Converted from GL model (log)
ShuffleNetV2b x0.5 41.41 19.07 1,366,792 42.37M Converted from GL model (log)
ShuffleNetV2c x0.5 40.31 18.51 1,366,792 42.37M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.0 34.26 13.54 2,278,604 147.92M Converted from GL model (log)
ShuffleNetV2c x1.0 30.98 11.61 2,279,760 148.85M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.5 32.82 12.69 4,406,098 318.61M Converted from GL model (log)
ShuffleNetV2 x2.0 32.45 12.49 7,601,686 593.66M Converted from GL model (log)
108-MENet-8x1 (g=3) 43.94 20.76 654,516 40.64M Converted from GL model (log)
128-MENet-8x1 (g=4) 42.43 19.59 750,796 43.58M Converted from GL model (log)
228-MENet-12x1 (g=3) 33.57 13.28 1,806,568 148.93M From clavichord93/MENet (log)
256-MENet-12x1 (g=4) 33.41 13.26 1,888,240 146.11M From clavichord93/MENet (log)
348-MENet-12x1 (g=3) 30.10 10.92 3,368,128 306.31M From clavichord93/MENet (log)
352-MENet-12x1 (g=8) 33.31 13.08 2,272,872 151.03M From clavichord93/MENet (log)
456-MENet-24x1 (g=3) 28.40 9.93 5,304,784 560.72M From clavichord93/MENet (log)
MobileNet x0.25 46.26 22.49 470,072 42.30M Converted from GL model (log)
MobileNet x0.5 36.30 15.14 1,331,592 152.04M Converted from GL model (log)
MobileNet x0.75 30.14 10.76 2,585,560 329.22M Converted from GL model (log)
MobileNet x1.0 29.86 10.36 4,231,976 573.83M From dmlc/gluon-cv (log)
FD-MobileNet x0.25 55.77 31.32 383,160 12.44M From clavichord93/FD-MobileNet (log)
FD-MobileNet x0.5 43.13 20.15 993,928 40.93M Converted from GL model (log)
FD-MobileNet x1.0 34.70 14.05 2,901,288 146.08M From clavichord93/FD-MobileNet (log)
MobileNetV2 x0.25 49.72 25.87 1,516,392 32.22M From dmlc/gluon-cv (log)
MobileNetV2 x0.5 36.54 15.19 1,964,736 95.62M From dmlc/gluon-cv (log)
MobileNetV2 x0.75 31.89 11.76 2,627,592 191.61M From dmlc/gluon-cv (log)
MobileNetV2 x1.0 29.31 10.39 3,504,960 320.19M From dmlc/gluon-cv (log)
IGCV3 28.40 9.84 3,491,688 323.81M From homles11/IGCV3 (log)
MnasNet 31.58 11.74 4,308,816 310.75M From zeusees/Mnasnet...Model (log)
DARTS 26.70 8.74 4,718,752 537.64M From quark0/darts (log)
Xception 20.97 5.49 22,855,952 8,385.86M From Cadene/pretrained...pytorch (log)
InceptionV3 21.12 5.65 23,834,568 5,746.72M From dmlc/gluon-cv (log)
InceptionV4 20.64 5.29 42,679,816 12,314.17M From Cadene/pretrained...pytorch (log)
InceptionResNetV2 19.93 4.90 55,843,464 13,189.58M From Cadene/pretrained...pytorch (log)
PolyNet 19.10 4.52 95,366,600 34,768.84M From Cadene/pretrained...pytorch (log)
NASNet-A 4@1056 25.68 8.16 5,289,978 587.29M From Cadene/pretrained...pytorch (log)
NASNet-A 6@4032 18.14 4.21 88,753,150 24,021.18M From Cadene/pretrained...pytorch (log)
PNASNet-5-Large 17.88 4.28 86,057,668 25,169.47M From Cadene/pretrained...pytorch (log)

For Chainer

Model Top1 Top5 Params FLOPs Remarks
AlexNet 44.08 21.32 61,100,840 715.49M From dmlc/gluon-cv (log)
VGG-11 31.89 11.79 132,863,336 7,622.65M From dmlc/gluon-cv (log)
VGG-13 31.06 11.16 133,047,848 11,326.85M From dmlc/gluon-cv (log)
VGG-16 26.75 8.70 138,357,544 15,489.95M From dmlc/gluon-cv (log)
VGG-19 25.86 8.23 143,667,240 19,653.05M From dmlc/gluon-cv (log)
BN-VGG-11b 30.37 10.60 132,868,840 7,622.65M From dmlc/gluon-cv (log)
BN-VGG-13b 29.45 10.19 133,053,736 11,326.85M From dmlc/gluon-cv (log)
BN-VGG-16b 26.89 8.63 138,365,992 15,489.95M From dmlc/gluon-cv (log)
BN-VGG-19b 25.65 8.16 143,678,248 19,653.05M From dmlc/gluon-cv (log)
ResNet-10 37.12 15.49 5,418,792 892.62M Converted from GL model (log)
ResNet-12 35.86 14.48 5,492,776 1,124.23M Converted from GL model (log)
ResNet-14 32.84 12.42 5,788,200 1,355.64M Converted from GL model (log)
ResNet-16 30.66 11.07 6,968,872 1,586.95M Converted from GL model (log)
ResNet-18 x0.25 49.08 24.48 831,096 136.64M Converted from GL model (log)
ResNet-18 x0.5 36.55 14.99 3,055,880 485.22M Converted from GL model (log)
ResNet-18 x0.75 33.27 12.56 6,675,352 1,045.75M Converted from GL model (log)
ResNet-18 29.08 9.97 11,689,512 1,818.21M Converted from GL model (log)
ResNet-34 25.35 7.95 21,797,672 3,669.16M From dmlc/gluon-cv (log)
ResNet-50 23.50 6.83 25,557,032 3,868.96M From dmlc/gluon-cv (log)
ResNet-50b 22.93 6.46 25,557,032 4,100.70M From dmlc/gluon-cv (log)
ResNet-101 21.65 6.01 44,549,160 7,586.30M From dmlc/gluon-cv (log)
ResNet-101b 21.16 5.59 44,549,160 7,818.04M From dmlc/gluon-cv (log)
ResNet-152 21.07 5.67 60,192,808 11,304.85M From dmlc/gluon-cv (log)
ResNet-152b 20.44 5.39 60,192,808 11,536.58M From dmlc/gluon-cv (log)
PreResNet-18 28.66 9.92 11,687,848 1,818.41M Converted from GL model (log)
PreResNet-34 25.89 8.12 21,796,008 3,669.36M From dmlc/gluon-cv (log)
PreResNet-50 23.36 6.69 25,549,480 3,869.16M From dmlc/gluon-cv (log)
PreResNet-50b 23.08 6.67 25,549,480 4,100.90M From dmlc/gluon-cv (log)
PreResNet-101 21.45 5.75 44,541,608 7,586.50M From dmlc/gluon-cv (log)
PreResNet-101b 21.61 5.87 44,541,608 7,818.24M From dmlc/gluon-cv (log)
PreResNet-152 20.73 5.30 60,185,256 11,305.05M From dmlc/gluon-cv (log)
PreResNet-152b 20.88 5.66 60,185,256 11,536.78M From dmlc/gluon-cv (log)
PreResNet-200b 21.03 5.60 64,666,280 15,040.27M From tornadomeet/ResNet (log)
ResNeXt-101 (32x4d) 21.11 5.69 44,177,704 7,991.62M From Cadene/pretrained...pytorch (log)
ResNeXt-101 (64x4d) 20.57 5.43 83,455,272 15,491.88M From Cadene/pretrained...pytorch (log)
SE-ResNet-50 22.53 6.41 28,088,024 3,877.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-101 21.90 5.88 49,326,872 7,600.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-152 21.46 5.77 66,821,848 11,324.62M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-50 (32x4d) 21.04 5.58 27,559,896 4,253.33M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-101 (32x4d) 19.99 5.01 48,955,416 8,005.33M From Cadene/pretrained...pytorch (log)
SENet-154 18.79 4.63 115,088,984 20,742.40M From Cadene/pretrained...pytorch (log)
AirNet50-1x64d (r=2) 22.46 6.20 27,425,864 4,757.77M From soeaver/AirNet-PyTorch (log)
AirNet50-1x64d (r=16) 22.89 6.50 25,714,952 4,385.54M From soeaver/AirNet-PyTorch (log)
AirNeXt50-32x4d (r=2) 21.50 5.73 27,604,296 5,321.18M From soeaver/AirNet-PyTorch (log)
BAM-ResNet-50 23.71 6.97 25,915,099 4,211.08M From Jongchan/attention-module (log)
CBAM-ResNet-50 22.99 6.40 28,089,624 4,118.19M From Jongchan/attention-module (log)
PyramidNet-101 (a=360) 22.66 6.49 42,455,070 8,706.81M From dyhan0920/Pyramid...PyTorch (log)
DiracNetV2-18 30.60 11.13 11,511,784 1,798.43M From szagoruyko/diracnets (log)
DiracNetV2-34 27.90 9.48 21,616,232 3,649.37M From szagoruyko/diracnets (log)
DenseNet-121 25.04 7.79 7,978,856 2,852.39M From dmlc/gluon-cv (log)
DenseNet-161 22.36 6.20 28,681,000 7,761.25M From dmlc/gluon-cv (log)
DenseNet-169 23.85 6.86 14,149,480 3,381.48M From dmlc/gluon-cv (log)
DenseNet-201 22.64 6.29 20,013,928 4,318.75M From dmlc/gluon-cv (log)
CondenseNet-74 (C=G=4) 26.81 8.61 4,773,944 533.64M From ShichenLiu/CondenseNet (log)
CondenseNet-74 (C=G=8) 29.74 10.43 2,935,416 278.55M From ShichenLiu/CondenseNet (log)
WRN-50-2 22.06 6.13 68,849,128 11,412.82M From szagoruyko/functional-zoo (log)
DRN-C-26 25.68 7.88 21,126,584 20,838.70M From fyu/drn (log)
DRN-C-42 23.72 6.93 31,234,744 31,236.97M From fyu/drn (log)
DRN-C-58 22.35 6.29 40,542,008 36,862.32M From fyu/drn (log)
DRN-D-22 26.65 8.50 16,393,752 16,626.00M From fyu/drn (log)
DRN-D-38 24.53 7.36 26,501,912 27,024.27M From fyu/drn (log)
DRN-D-54 22.08 6.23 35,809,176 32,649.62M From fyu/drn (log)
DRN-D-105 21.32 5.84 54,801,304 48,682.11M From fyu/drn (log)
DPN-68 23.61 7.01 12,611,602 2,338.71M From Cadene/pretrained...pytorch (log)
DPN-98 20.80 5.53 61,570,728 11,702.80M From Cadene/pretrained...pytorch (log)
DPN-131 20.04 5.23 79,254,504 16,056.22M From Cadene/pretrained...pytorch (log)
DarkNet Tiny 40.33 17.46 1,042,104 496.34M Converted from GL model (log)
DarkNet Ref 38.09 16.71 7,319,416 365.55M Converted from GL model (log)
SqueezeNet v1.0 38.76 17.38 1,248,424 828.30M Converted from GL model (log)
SqueezeNet v1.1 39.13 17.40 1,235,496 354.88M Converted from GL model (log)
SqueezeResNet v1.1 39.85 17.87 1,235,496 354.88M Converted from GL model (log)
ShuffleNetV2 x0.5 43.45 20.73 1,366,792 42.34M Converted from GL model (log)
ShuffleNetV2b x0.5 40.95 18.56 1,366,792 42.37M Converted from GL model (log)
ShuffleNetV2c x0.5 39.82 18.14 1,366,792 42.37M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.0 35.69 14.71 2,278,604 147.92M Converted from GL model (log)
ShuffleNetV2c x1.0 30.74 11.37 2,279,760 148.85M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.5 33.96 13.37 4,406,098 318.61M Converted from GL model (log)
ShuffleNetV2 x2.0 33.21 13.03 7,601,686 593.66M Converted from GL model (log)
108-MENet-8x1 (g=3) 43.67 20.42 654,516 40.64M Converted from GL model (log)
128-MENet-8x1 (g=4) 42.07 19.19 750,796 43.58M Converted from GL model (log)
228-MENet-12x1 (g=3) 34.93 14.01 1,806,568 148.93M From clavichord93/MENet (log)
256-MENet-12x1 (g=4) 34.44 13.91 1,888,240 146.11M From clavichord93/MENet (log)
348-MENet-12x1 (g=3) 31.14 11.40 3,368,128 306.31M From clavichord93/MENet (log)
352-MENet-12x1 (g=8) 34.62 13.68 2,272,872 151.03M From clavichord93/MENet (log)
456-MENet-24x1 (g=3) 29.55 10.39 5,304,784 560.72M From clavichord93/MENet (log)
MobileNet x0.25 45.85 22.16 470,072 42.30M Converted from GL model (log)
MobileNet x0.5 36.15 14.86 1,331,592 152.04M Converted from GL model (log)
MobileNet x0.75 29.86 10.53 2,585,560 329.22M Converted from GL model (log)
MobileNet x1.0 29.71 10.31 4,231,976 573.83M From dmlc/gluon-cv (log)
FD-MobileNet x0.25 56.11 31.45 383,160 12.44M Converted from GL model (log)
FD-MobileNet x0.5 42.68 19.76 993,928 40.93M Converted from GL model (log)
FD-MobileNet x1.0 34.44 13.74 2,901,288 146.08M Converted from GL model (log)
MobileNetV2 x0.25 49.11 25.49 1,516,392 32.22M From dmlc/gluon-cv (log)
MobileNetV2 x0.5 35.96 14.98 1,964,736 95.62M From dmlc/gluon-cv (log)
MobileNetV2 x0.75 31.28 11.48 2,627,592 191.61M From dmlc/gluon-cv (log)
MobileNetV2 x1.0 28.87 10.05 3,504,960 320.19M From dmlc/gluon-cv (log)
IGCV3 28.20 9.55 3,491,688 323.81M From homles11/IGCV3 (log)
MnasNet 31.27 11.44 4,308,816 310.75M From zeusees/Mnasnet...Model (log)
DARTS 27.29 8.97 4,718,752 537.64M From quark0/darts (log)
Xception 21.04 5.47 22,855,952 8,385.86M From Cadene/pretrained...pytorch (log)
InceptionV3 21.11 5.61 23,834,568 5,746.72M From dmlc/gluon-cv (log)
InceptionV4 20.62 5.26 42,679,816 12,314.17M From Cadene/pretrained...pytorch (log)
InceptionResNetV2 19.93 4.92 55,843,464 13,189.58M From Cadene/pretrained...pytorch (log)
PolyNet 19.08 4.50 95,366,600 34,768.84M From Cadene/pretrained...pytorch (log)
NASNet-A 4@1056 25.36 7.96 5,289,978 587.29M From Cadene/pretrained...pytorch (log)
NASNet-A 6@4032 18.17 4.22 88,753,150 24,021.18M From Cadene/pretrained...pytorch (log)
PNASNet-5-Large 17.90 4.26 86,057,668 25,169.47M From Cadene/pretrained...pytorch (log)

For Keras

Model Top1 Top5 Params FLOPs Remarks
AlexNet 44.10 21.26 61,100,840 715.49M From dmlc/gluon-cv (log)
VGG-11 31.90 11.75 132,863,336 7,622.65M From dmlc/gluon-cv (log)
VGG-13 31.06 11.12 133,047,848 11,326.85M From dmlc/gluon-cv (log)
VGG-16 26.78 8.69 138,357,544 15,489.95M From dmlc/gluon-cv (log)
VGG-19 25.87 8.23 143,667,240 19,653.05M From dmlc/gluon-cv (log)
BN-VGG-11b 30.34 10.57 132,868,840 7,622.65M From dmlc/gluon-cv (log)
BN-VGG-13b 29.48 10.16 133,053,736 11,326.85M From dmlc/gluon-cv (log)
BN-VGG-16b 26.88 8.65 138,365,992 15,489.95M From dmlc/gluon-cv (log)
BN-VGG-19b 25.65 8.14 143,678,248 19,653.05M From dmlc/gluon-cv (log)
ResNet-10 37.09 15.54 5,418,792 892.62M Converted from GL model (log)
ResNet-12 35.86 14.45 5,492,776 1,124.23M Converted from GL model (log)
ResNet-14 32.85 12.42 5,788,200 1,355.64M Converted from GL model (log)
ResNet-16 30.67 11.09 6,968,872 1,586.95M Converted from GL model (log)
ResNet-18 x0.25 49.14 24.45 831,096 136.64M Converted from GL model (log)
ResNet-18 x0.5 36.54 14.96 3,055,880 485.22M Converted from GL model (log)
ResNet-18 x0.75 33.24 12.54 6,675,352 1,045.75M Converted from GL model (log)
ResNet-18 29.13 9.94 11,689,512 1,818.21M Converted from GL model (log)
ResNet-34 25.32 7.92 21,797,672 3,669.16M From dmlc/gluon-cv (log)
ResNet-50 23.49 6.87 25,557,032 3,868.96M From dmlc/gluon-cv (log)
ResNet-50b 22.90 6.44 25,557,032 4,100.70M From dmlc/gluon-cv (log)
ResNet-101 21.64 5.99 44,549,160 7,586.30M From dmlc/gluon-cv (log)
ResNet-101b 21.17 5.60 44,549,160 7,818.04M From dmlc/gluon-cv (log)
ResNet-152 21.00 5.61 60,192,808 11,304.85M From dmlc/gluon-cv (log)
ResNet-152b 20.53 5.37 60,192,808 11,536.58M From dmlc/gluon-cv (log)
PreResNet-18 28.72 9.88 11,687,848 1,818.41M Converted from GL model (log)
PreResNet-34 25.86 8.11 21,796,008 3,669.36M From dmlc/gluon-cv (log)
PreResNet-50 23.38 6.68 25,549,480 3,869.16M From dmlc/gluon-cv (log)
PreResNet-50b 23.14 6.63 25,549,480 4,100.90M From dmlc/gluon-cv (log)
PreResNet-101 21.43 5.75 44,541,608 7,586.50M From dmlc/gluon-cv (log)
PreResNet-101b 21.71 5.88 44,541,608 7,818.24M From dmlc/gluon-cv (log)
PreResNet-152 20.69 5.31 60,185,256 11,305.05M From dmlc/gluon-cv (log)
PreResNet-152b 20.99 5.76 60,185,256 11,536.78M From dmlc/gluon-cv (log)
PreResNet-200b 21.09 5.64 64,666,280 15,040.27M From tornadomeet/ResNet (log)
ResNeXt-101 (32x4d) 21.30 5.78 44,177,704 7,991.62M From Cadene/pretrained...pytorch (log)
ResNeXt-101 (64x4d) 20.59 5.41 83,455,272 15,491.88M From Cadene/pretrained...pytorch (log)
SE-ResNet-50 22.50 6.43 28,088,024 3,877.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-101 21.92 5.88 49,326,872 7,600.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-152 21.46 5.77 66,821,848 11,324.62M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-50 (32x4d) 21.05 5.57 27,559,896 4,253.33M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-101 (32x4d) 19.98 4.99 48,955,416 8,005.33M From Cadene/pretrained...pytorch (log)
SENet-154 18.83 4.65 115,088,984 20,742.40M From Cadene/pretrained...pytorch (log)
DenseNet-121 25.09 7.80 7,978,856 2,852.39M From dmlc/gluon-cv (log)
DenseNet-161 22.39 6.18 28,681,000 7,761.25M From dmlc/gluon-cv (log)
DenseNet-169 23.88 6.89 14,149,480 3,381.48M From dmlc/gluon-cv (log)
DenseNet-201 22.69 6.35 20,013,928 4,318.75M From dmlc/gluon-cv (log)
DarkNet Tiny 40.31 17.46 1,042,104 496.34M Converted from GL model (log)
DarkNet Ref 37.99 16.68 7,319,416 365.55M Converted from GL model (log)
SqueezeNet v1.0 39.17 17.56 1,248,424 828.30M Converted from GL model (log)
SqueezeNet v1.1 39.08 17.39 1,235,496 354.88M Converted from GL model (log)
SqueezeResNet v1.1 39.82 17.84 1,235,496 354.88M Converted from GL model (log)
ShuffleNetV2 x0.5 40.76 18.40 1,366,792 42.34M Converted from GL model (log)
ShuffleNetV2 x1.0 33.79 13.38 2,278,604 147.92M Converted from GL model (log)
ShuffleNetV2 x1.5 32.46 12.47 4,406,098 318.61M Converted from GL model (log)
ShuffleNetV2 x2.0 31.91 12.23 7,601,686 593.66M Converted from GL model (log)
108-MENet-8x1 (g=3) 43.61 20.31 654,516 40.64M Converted from GL model (log)
128-MENet-8x1 (g=4) 42.08 19.14 750,796 43.58M Converted from GL model (log)
228-MENet-12x1 (g=3) 35.02 14.01 1,806,568 148.93M From clavichord93/MENet (log)
256-MENet-12x1 (g=4) 34.48 13.91 1,888,240 146.11M From clavichord93/MENet (log)
348-MENet-12x1 (g=3) 31.17 11.42 3,368,128 306.31M From clavichord93/MENet (log)
352-MENet-12x1 (g=8) 34.69 13.75 2,272,872 151.03M From clavichord93/MENet (log)
456-MENet-24x1 (g=3) 29.55 10.44 5,304,784 560.72M From clavichord93/MENet (log)
MobileNet x0.25 45.80 22.17 470,072 42.30M Converted from GL model (log)
MobileNet x0.5 36.11 14.81 1,331,592 152.04M Converted from GL model (log)
MobileNet x0.75 29.85 10.51 2,585,560 329.22M Converted from GL model (log)
MobileNet x1.0 29.24 10.03 4,231,976 573.83M From dmlc/gluon-cv (log)
FD-MobileNet x0.25 56.17 31.37 383,160 12.44M Converted from GL model (log)
FD-MobileNet x0.5 42.61 19.69 993,928 40.93M Converted from GL model (log)
FD-MobileNet x1.0 34.42 13.74 2,901,288 146.08M Converted from GL model (log)
MobileNetV2 x0.25 48.86 25.24 1,516,392 32.22M From dmlc/gluon-cv (log)
MobileNetV2 x0.5 35.51 14.65 1,964,736 95.62M From dmlc/gluon-cv (log)
MobileNetV2 x0.75 30.81 11.26 2,627,592 191.61M From dmlc/gluon-cv (log)
MobileNetV2 x1.0 28.50 9.90 3,504,960 320.19M From dmlc/gluon-cv (log)
IGCV3 28.21 9.55 3,491,688 323.81M From homles11/IGCV3 (log)
MnasNet 31.30 11.45 4,308,816 310.75M From zeusees/Mnasnet...Model (log)

For TensorFlow

Model Top1 Top5 Params FLOPs Remarks
AlexNet 44.07 21.32 61,100,840 715.49M From dmlc/gluon-cv (log)
VGG-11 31.89 11.73 132,863,336 7,622.65M From dmlc/gluon-cv (log)
VGG-13 31.03 11.15 133,047,848 11,326.85M From dmlc/gluon-cv (log)
VGG-16 26.77 8.68 138,357,544 15,489.95M From dmlc/gluon-cv (log)
VGG-19 25.93 8.23 143,667,240 19,653.05M From dmlc/gluon-cv (log)
BN-VGG-11b 30.34 10.58 132,868,840 7,622.65M From dmlc/gluon-cv (log)
BN-VGG-13b 29.47 10.15 133,053,736 11,326.85M From dmlc/gluon-cv (log)
BN-VGG-16b 26.83 8.66 138,365,992 15,489.95M From dmlc/gluon-cv (log)
BN-VGG-19b 25.62 8.17 143,678,248 19,653.05M From dmlc/gluon-cv (log)
ResNet-10 37.11 15.52 5,418,792 892.62M Converted from GL model (log)
ResNet-12 35.82 14.50 5,492,776 1,124.23M Converted from GL model (log)
ResNet-14 32.83 12.45 5,788,200 1,355.64M Converted from GL model (log)
ResNet-16 30.66 11.05 6,968,872 1,586.95M Converted from GL model (log)
ResNet-18 x0.25 49.12 24.50 831,096 136.64M Converted from GL model (log)
ResNet-18 x0.5 36.51 14.93 3,055,880 485.22M Converted from GL model (log)
ResNet-18 x0.75 33.28 12.50 6,675,352 1,045.75M Converted from GL model (log)
ResNet-18 29.10 9.99 11,689,512 1,818.21M Converted from GL model (log)
ResNet-34 25.32 7.93 21,797,672 3,669.16M From dmlc/gluon-cv (log)
ResNet-50 23.48 6.87 25,557,032 3,868.96M From dmlc/gluon-cv (log)
ResNet-50b 22.97 6.48 25,557,032 4,100.70M From dmlc/gluon-cv (log)
ResNet-101 21.61 6.01 44,549,160 7,586.30M From dmlc/gluon-cv (log)
ResNet-101b 21.22 5.57 44,549,160 7,818.04M From dmlc/gluon-cv (log)
ResNet-152 20.99 5.59 60,192,808 11,304.85M From dmlc/gluon-cv (log)
ResNet-152b 20.55 5.35 60,192,808 11,536.58M From dmlc/gluon-cv (log)
PreResNet-18 28.75 9.88 11,687,848 1,818.41M Converted from GL model (log)
PreResNet-34 25.82 8.08 21,796,008 3,669.36M From dmlc/gluon-cv (log)
PreResNet-50 23.42 6.68 25,549,480 3,869.16M From dmlc/gluon-cv (log)
PreResNet-50b 23.12 6.61 25,549,480 4,100.90M From dmlc/gluon-cv (log)
PreResNet-101 21.49 5.72 44,541,608 7,586.50M From dmlc/gluon-cv (log)
PreResNet-101b 21.70 5.91 44,541,608 7,818.24M From dmlc/gluon-cv (log)
PreResNet-152 20.63 5.29 60,185,256 11,305.05M From dmlc/gluon-cv (log)
PreResNet-152b 20.95 5.76 60,185,256 11,536.78M From dmlc/gluon-cv (log)
PreResNet-200b 21.12 5.60 64,666,280 15,040.27M From tornadomeet/ResNet (log)
ResNeXt-101 (32x4d) 21.33 5.80 44,177,704 7,991.62M From Cadene/pretrained...pytorch (log)
ResNeXt-101 (64x4d) 20.59 5.43 83,455,272 15,491.88M From Cadene/pretrained...pytorch (log)
SE-ResNet-50 22.53 6.43 28,088,024 3,877.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-101 21.92 5.89 49,326,872 7,600.01M From Cadene/pretrained...pytorch (log)
SE-ResNet-152 21.48 5.78 66,821,848 11,324.62M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-50 (32x4d) 21.01 5.53 27,559,896 4,253.33M From Cadene/pretrained...pytorch (log)
SE-ResNeXt-101 (32x4d) 19.99 4.97 48,955,416 8,005.33M From Cadene/pretrained...pytorch (log)
SENet-154 18.77 4.63 115,088,984 20,742.40M From Cadene/pretrained...pytorch (log)
DenseNet-121 25.16 7.82 7,978,856 2,852.39M From dmlc/gluon-cv (log)
DenseNet-161 22.40 6.17 28,681,000 7,761.25M From dmlc/gluon-cv (log)
DenseNet-169 23.93 6.87 14,149,480 3,381.48M From dmlc/gluon-cv (log)
DenseNet-201 22.70 6.35 20,013,928 4,318.75M From dmlc/gluon-cv (log)
DarkNet Tiny 40.35 17.51 1,042,104 496.34M Converted from GL model (log)
DarkNet Ref 37.99 16.72 7,319,416 365.55M Converted from GL model (log)
SqueezeNet v1.0 39.18 17.58 1,248,424 828.30M Converted from GL model (log)
SqueezeNet v1.1 39.14 17.39 1,235,496 354.88M Converted from GL model (log)
SqueezeResNet v1.1 39.75 17.92 1,235,496 354.88M Converted from GL model (log)
ShuffleNetV2 x0.5 40.88 18.44 1,366,792 42.34M Converted from GL model (log)
ShuffleNetV2b x0.5 41.03 18.59 1,366,792 42.37M Converted from GL model (log)
ShuffleNetV2c x0.5 39.93 18.11 1,366,792 42.37M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.0 33.81 13.40 2,278,604 147.92M Converted from GL model (log)
ShuffleNetV2c x1.0 30.77 11.39 2,279,760 148.85M From tensorpack/tensorpack (log)
ShuffleNetV2 x1.5 32.51 12.50 4,406,098 318.61M Converted from GL model (log)
ShuffleNetV2 x2.0 31.99 12.26 7,601,686 593.66M Converted from GL model (log)
108-MENet-8x1 (g=3) 43.67 20.32 654,516 40.64M Converted from GL model (log)
128-MENet-8x1 (g=4) 42.04 19.15 750,796 43.58M Converted from GL model (log)
228-MENet-12x1 (g=3) 35.01 14.05 1,806,568 148.93M From clavichord93/MENet (log)
256-MENet-12x1 (g=4) 34.48 13.95 1,888,240 146.11M From clavichord93/MENet (log)
348-MENet-12x1 (g=3) 31.19 11.41 3,368,128 306.31M From clavichord93/MENet (log)
352-MENet-12x1 (g=8) 34.65 13.71 2,272,872 151.03M From clavichord93/MENet (log)
456-MENet-24x1 (g=3) 29.56 10.46 5,304,784 560.72M From clavichord93/MENet (log)
MobileNet x0.25 45.78 22.21 470,072 42.30M Converted from GL model (log)
MobileNet x0.5 36.18 14.84 1,331,592 152.04M Converted from GL model (log)
MobileNet x0.75 29.82 10.49 2,585,560 329.22M Converted from GL model (log)
MobileNet x1.0 29.30 10.04 4,231,976 573.83M From dmlc/gluon-cv (log)
FD-MobileNet x0.25 56.08 31.44 383,160 12.44M Converted from GL model (log)
FD-MobileNet x0.5 42.67 19.70 993,928 40.93M Converted from GL model (log)
FD-MobileNet x1.0 34.47 13.74 2,901,288 146.08M Converted from GL model (log)
MobileNetV2 x0.25 48.87 25.26 1,516,392 32.22M From dmlc/gluon-cv (log)
MobileNetV2 x0.5 35.51 14.60 1,964,736 95.62M From dmlc/gluon-cv (log)
MobileNetV2 x0.75 30.79 11.24 2,627,592 191.61M From dmlc/gluon-cv (log)
MobileNetV2 x1.0 28.53 9.90 3,504,960 320.19M From dmlc/gluon-cv (log)
IGCV3 28.17 9.55 3,491,688 323.81M From homles11/IGCV3 (log)
MnasNet 31.29 11.44 4,308,816 310.75M From zeusees/Mnasnet...Model (log)