#!/usr/bin/env python
# ******************************************************************************
# Copyright 2021 Brainchip Holdings Ltd.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ******************************************************************************
"""
AkidaNet model definition for ImageNet classification.
AkidaNet is an NSoC optimized model inspired from VGG and MobileNet V1
architectures. It can be used for multiple use cases through transfer learning.
"""
from keras import Model, regularizers
from keras.layers import (Input, Reshape, Dropout, Activation, Rescaling)
from cnn2snn import quantize, load_quantized_model
from .imagenet_utils import obtain_input_shape
from ..layer_blocks import conv_block, separable_conv_block
from ..utils import fetch_file
BASE_WEIGHT_PATH = 'http://data.brainchip.com/models/akidanet/'
[docs]def akidanet_imagenet(input_shape=None,
alpha=1.0,
include_top=True,
pooling=None,
classes=1000,
weight_quantization=0,
activ_quantization=0,
input_weight_quantization=None,
input_scaling=(128, -1)):
"""Instantiates the AkidaNet architecture.
Args:
input_shape (tuple, optional): shape tuple. Defaults to None.
alpha (float, optional): controls the width of the model.
Defaults to 1.0.
* If `alpha` < 1.0, proportionally decreases the number of filters
in each layer.
* If `alpha` > 1.0, proportionally increases the number of filters
in each layer.
* If `alpha` = 1, default number of filters from the paper are used
at each layer.
include_top (bool, optional): whether to include the fully-connected
layer at the top of the model. Defaults to True.
pooling (str, optional): optional pooling mode for feature extraction
when `include_top` is `False`.
Defaults to None.
* `None` means that the output of the model will be the 4D tensor
output of the last convolutional block.
* `avg` means that global average pooling will be applied to the
output of the last convolutional block, and thus the output of the
model will be a 2D tensor.
classes (int, optional): optional number of classes to classify images
into, only to be specified if `include_top` is `True`. Defaults to 1000.
weight_quantization (int, optional): sets all weights in the model to have
a particular quantization bitwidth except for the weights in the
first layer.
Defaults to 0.
* '0' implements floating point 32-bit weights.
* '2' through '8' implements n-bit weights where n is from 2-8 bits.
activ_quantization (int, optional): sets all activations in the model to have a
particular activation quantization bitwidth.
Defaults to 0.
* '0' implements floating point 32-bit activations.
* '2' through '8' implements n-bit weights where n is from 2-8 bits.
input_weight_quantization (int, optional): sets weight quantization in the first layer.
Defaults to weight_quantization value.
* '0' implements floating point 32-bit weights.
* '2' through '8' implements n-bit weights where n is from 2-8 bits.
input_scaling (tuple, optional): scale factor and offset to apply to
inputs. Defaults to (128, -1). Note that following Akida convention,
the scale factor is an integer used as a divider.
Returns:
keras.Model: a Keras model for AkidaNet/ImageNet.
Raises:
ValueError: in case of invalid input shape.
"""
# check if overrides have been provided and override
if input_weight_quantization is None:
input_weight_quantization = weight_quantization
# Define weight regularization, will apply to the convolutional layers and
# to all pointwise weights of separable convolutional layers.
weight_regularizer = regularizers.l2(4e-5)
# Determine proper input shape and default size.
if input_shape is None:
default_size = 224
else:
rows = input_shape[0]
cols = input_shape[1]
if rows == cols and rows in [128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
input_shape = obtain_input_shape(input_shape,
default_size=default_size,
min_size=32,
include_top=include_top)
img_input = Input(shape=input_shape, name="input")
if input_scaling is None:
x = img_input
else:
scale, offset = input_scaling
x = Rescaling(1. / scale, offset, name="rescaling")(img_input)
x = conv_block(x,
filters=int(32 * alpha),
name='conv_0',
kernel_size=(3, 3),
padding='same',
use_bias=False,
strides=2,
add_batchnorm=True,
add_activation=True,
kernel_regularizer=weight_regularizer)
x = conv_block(x,
filters=int(64 * alpha),
name='conv_1',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
kernel_regularizer=weight_regularizer)
x = conv_block(x,
filters=int(128 * alpha),
name='conv_2',
kernel_size=(3, 3),
padding='same',
strides=2,
use_bias=False,
add_batchnorm=True,
add_activation=True,
kernel_regularizer=weight_regularizer)
x = conv_block(x,
filters=int(128 * alpha),
name='conv_3',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
kernel_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(256 * alpha),
name='separable_4',
kernel_size=(3, 3),
padding='same',
strides=2,
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(256 * alpha),
name='separable_5',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_6',
kernel_size=(3, 3),
padding='same',
strides=2,
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_7',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_8',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_9',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_10',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(512 * alpha),
name='separable_11',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
x = separable_conv_block(x,
filters=int(1024 * alpha),
name='separable_12',
kernel_size=(3, 3),
padding='same',
strides=2,
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
# Last separable layer with global pooling
layer_pooling = 'global_avg' if include_top or pooling == 'avg' else None
x = separable_conv_block(x,
filters=int(1024 * alpha),
name='separable_13',
kernel_size=(3, 3),
padding='same',
pooling=layer_pooling,
use_bias=False,
add_batchnorm=True,
add_activation=True,
pointwise_regularizer=weight_regularizer)
if include_top:
shape = (1, 1, int(1024 * alpha))
x = Reshape(shape, name='reshape_1')(x)
x = Dropout(1e-3, name='dropout')(x)
x = separable_conv_block(x,
filters=classes,
name='separable_14',
kernel_size=(3, 3),
padding='same',
use_bias=False,
add_batchnorm=False,
add_activation=False,
pointwise_regularizer=weight_regularizer)
act_function = 'softmax' if classes > 1 else 'sigmoid'
x = Activation(act_function, name=f'act_{act_function}')(x)
x = Reshape((classes,), name='reshape_2')(x)
# Create model.
model = Model(img_input,
x,
name='akidanet_%0.2f_%s_%s' %
(alpha, input_shape[0], classes))
if ((weight_quantization != 0) or (activ_quantization != 0) or
(input_weight_quantization != 0)):
return quantize(model, weight_quantization, activ_quantization,
input_weight_quantization)
return model
[docs]def akidanet_imagenet_pretrained(alpha=1.0):
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
ImageNet dataset.
Args:
alpha (float, optional): width of the model, allowed values in [0.25,
0.5, 1]. Defaults to 1.0.
Returns:
keras.Model: a Keras Model instance.
"""
if alpha == 1.0:
model_name = 'akidanet_imagenet_224_iq8_wq4_aq4.h5'
file_hash = '88e2e77e9c1379f2972f269f04656fc28d29e4971cc50dc2c75b49c722a8d581'
elif alpha == 0.5:
model_name = 'akidanet_imagenet_224_alpha_50_iq8_wq4_aq4.h5'
file_hash = '609dd98260f9304b7a9f0efa6553d0b9789b8cf858dbaa962756028715ee238f'
elif alpha == 0.25:
model_name = 'akidanet_imagenet_224_alpha_25_iq8_wq4_aq4.h5'
file_hash = '4760ffd387fe08ce9aa28a8dcd30cf8c23af151855d7a07397c3dd3c0ad43fcc'
else:
raise ValueError(
f"Requested model with alpha={alpha} is not available.")
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_cats_vs_dogs_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
Cats vs.Dogs dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_cats_vs_dogs_iq8_wq4_aq4_lwq_float.h5'
file_hash = 'c3483a261bcbcdb5175c43bcf839bcd9124e2fb81c0a84ec4d84702f111313b2'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_imagenette_pretrained(alpha=1.0):
"""
Helper method to retrieve a `akidanet_imagenet` model that was trained on
Imagenette dataset.
Args:
alpha (float, optional): width of the model, allowed values in [0.25,
0.5, 1]. Defaults to 1.0.
Returns:
keras.Model: a Keras Model instance.
"""
if alpha == 1.0:
model_name = 'akidanet_imagenette_224_iq8_wq4_aq4.h5'
file_hash = 'e48c3083f529d3cb91db090e8bfb9b2669dfce2b4407391ca848f005e656b4d8'
elif alpha == 0.5:
model_name = 'akidanet_imagenette_224_alpha_50_iq8_wq4_aq4.h5'
file_hash = 'a96cf3c49dec0c3d4d6db6e964748402faa16565e174e785128dcdec3555a830'
elif alpha == 0.25:
model_name = 'akidanet_imagenette_224_alpha_25_iq8_wq4_aq4.h5'
file_hash = '90e2eb651cf490129c1e619017dd47b1e87367aeea0a5401da62782d8f33c19f'
else:
raise ValueError(
f"Requested model with alpha={alpha} is not available.")
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_faceidentification_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
CASIA Webface dataset and that performs face identification.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_faceidentification_iq8_wq4_aq4.h5'
file_hash = '117df1118124c5d2cc0cf867dcc10fa1dfbc3978c1bfe0dcba727ac369e48765'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_faceverification_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
CASIA Webface dataset and optimized with ArcFace that can perform face
verification on LFW.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_faceverification_iq8_wq4_aq4.h5'
file_hash = 'b52cba91acbda689587f103bfd28dff59c6b8cd10b4b1bbb09f0df96b37dabfb'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_melanoma_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
SIIM-ISIC Melanoma Classification dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_melanoma_iq8_wq4_aq4.h5'
file_hash = '98f5bbdd35089ef0ade17d3e1d7f80ed31f28bdf736ef5d4c4519adac813ac69'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_odir5k_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
ODIR-5K dataset.
The model focuses on the following classes that are a part of the original
dataset: normal, cataract, AMD (age related macular degeneration) and
pathological myopia.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_odir5k_iq8_wq4_aq4.h5'
file_hash = 'f7bd4b8a7d808541e4dcd7e078a3e9fada83e756b56513af89abb5c453c216c3'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_retinal_oct_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
retinal OCT dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_retinal_oct_iq8_wq4_aq4.h5'
file_hash = 'a5de3d3b6d5545870de2275ea59df7940e22f9eec863f4302e8600177ece9b58'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_ecg_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
ECG classification Physionet2017 dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_ecg_iq8_wq4_aq4.h5'
file_hash = 'a124f4a9ab8a02bab73714e457ef470a542afadf6ab4cd6cc52f10e6bee025e9'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_plantvillage_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
PlantVillage dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_plantvillage_iq8_wq4_aq4.h5'
file_hash = '1400910c774fd78a5e6ea227ff28cb28e79ecec0909a378068cd9f40ddaf4e0a'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_cifar10_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
CIFAR-10 dataset. Since CIFAR-10 images have a 32x32 size, they need to be
resized to match akidanet input layer. This can be done by calling the
'resize_image' function available under akida_models.cifar10.preprocessing.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_cifar10_iq8_wq4_aq4.h5'
file_hash = 'cf5bba0e0704af3ef9752910e8a0dfddaf96666fd4d724b2f31a8733f1e153c3'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)
[docs]def akidanet_vww_pretrained():
"""
Helper method to retrieve an `akidanet_imagenet` model that was trained on
VWW dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name = 'akidanet_vww_iq8_wq4_aq4.h5'
file_hash = 'b962e022071fcc77cb9403e1343341f3c160c870ba7d1f4a5011da67e4d33e66'
model_path = fetch_file(BASE_WEIGHT_PATH + model_name,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_quantized_model(model_path)