#!/usr/bin/env python
# ******************************************************************************
# Copyright 2020 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.
# ******************************************************************************
"""
VGG model definition for UTKFace regression.
"""
from keras import Model
from keras.layers import Dropout, Flatten, Input, Rescaling
from ..layer_blocks import conv_block, dense_block
from ..utils import fetch_file, get_params_by_version
from ..model_io import load_model, get_model_path
[docs]def vgg_utk_face(input_shape=(32, 32, 3), input_scaling=(127, -1)):
"""Instantiates a VGG-like model for the regression example on age
estimation using UTKFace dataset.
Args:
input_shape (tuple, optional): input shape tuple of the model. Defaults to (32, 32, 3).
input_scaling (tuple, optional): scale factor and offset to apply to
inputs. Defaults to (127, -1). Note that following Akida convention,
the scale factor is an integer used as a divider.
Returns:
keras.Model: a Keras model for VGG/UTKFace
"""
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)
# Model version management
_, _, relu_activation = get_params_by_version()
x = conv_block(x,
filters=32,
kernel_size=(3, 3),
name='conv_0',
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = conv_block(x,
filters=32,
kernel_size=(3, 3),
name='conv_1',
padding='same',
pooling='max',
pool_size=2,
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = Dropout(0.3, name="dropout_3")(x)
x = conv_block(x,
filters=64,
kernel_size=(3, 3),
padding='same',
name='conv_2',
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = conv_block(x,
filters=64,
kernel_size=(3, 3),
padding='same',
name='conv_3',
pooling='max',
pool_size=2,
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = Dropout(0.3, name="dropout_4")(x)
x = conv_block(x,
filters=84,
kernel_size=(3, 3),
padding='same',
name='conv_4',
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = Dropout(0.3, name="dropout_5")(x)
x = Flatten(name="flatten")(x)
x = dense_block(x,
units=64,
name='dense_1',
use_bias=False,
relu_activation=relu_activation,
add_batchnorm=True)
x = dense_block(x, units=1, name='dense_2', relu_activation=False)
return Model(img_input, x, name='vgg_utk_face')
[docs]def vgg_utk_face_pretrained():
"""
Helper method to retrieve a `vgg_utk_face` model that was trained on
UTK Face dataset.
Returns:
keras.Model: a Keras Model instance.
"""
model_name_v1 = 'vgg_utk_face_iq8_wq2_aq2.h5'
file_hash_v1 = 'e341d2d5e4655846ddc7aceff0d4e324cbfbcca16f3cfefc65e7b0863e4a23a3'
model_name_v2 = 'vgg_utk_face_i8_w4_a4.h5'
file_hash_v2 = '06a892b9e831767537b1da937bbacf05825c7493b7a5a9ba46bea4900de0563a'
model_path, model_name, file_hash = get_model_path("vgg", model_name_v1, file_hash_v1,
model_name_v2, file_hash_v2)
model_path = fetch_file(model_path,
fname=model_name,
file_hash=file_hash,
cache_subdir='models')
return load_model(model_path)