You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
PaddleSpeech/paddlespeech/vector/modules/loss.py

397 lines
15 KiB

# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
#
# 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.
# This is modified from SpeechBrain
# https://github.com/speechbrain/speechbrain/blob/085be635c07f16d42cd1295045bc46c407f1e15b/speechbrain/nnet/losses.py
import math
import paddle
import paddle.nn as nn
import paddle.nn.functional as F
from paddle.nn import initializer as I
class AngularMargin(nn.Layer):
def __init__(self, margin=0.0, scale=1.0):
"""An implementation of Angular Margin (AM) proposed in the following
paper: '''Margin Matters: Towards More Discriminative Deep Neural Network
Embeddings for Speaker Recognition''' (https://arxiv.org/abs/1906.07317)
Args:
margin (float, optional): The margin for cosine similiarity. Defaults to 0.0.
scale (float, optional): The scale for cosine similiarity. Defaults to 1.0.
"""
super(AngularMargin, self).__init__()
self.margin = margin
self.scale = scale
def forward(self, outputs, targets):
outputs = outputs - self.margin * targets
return self.scale * outputs
class AdditiveAngularMargin(AngularMargin):
def __init__(self, margin=0.0, scale=1.0, easy_margin=False):
"""The Implementation of Additive Angular Margin (AAM) proposed
in the following paper: '''Margin Matters: Towards More Discriminative Deep Neural Network Embeddings for Speaker Recognition'''
(https://arxiv.org/abs/1906.07317)
Args:
margin (float, optional): margin factor. Defaults to 0.0.
scale (float, optional): scale factor. Defaults to 1.0.
easy_margin (bool, optional): easy_margin flag. Defaults to False.
"""
super(AdditiveAngularMargin, self).__init__(margin, scale)
self.easy_margin = easy_margin
self.cos_m = math.cos(self.margin)
self.sin_m = math.sin(self.margin)
self.th = math.cos(math.pi - self.margin)
self.mm = math.sin(math.pi - self.margin) * self.margin
def forward(self, outputs, targets):
cosine = outputs.astype('float32')
sine = paddle.sqrt(1.0 - paddle.pow(cosine, 2))
phi = cosine * self.cos_m - sine * self.sin_m # cos(theta + m)
if self.easy_margin:
phi = paddle.where(cosine > 0, phi, cosine)
else:
phi = paddle.where(cosine > self.th, phi, cosine - self.mm)
outputs = (targets * phi) + ((1.0 - targets) * cosine)
return self.scale * outputs
class LogSoftmaxWrapper(nn.Layer):
def __init__(self, loss_fn):
"""Speaker identificatin loss function wrapper
including all of compositions of the loss transformation
Args:
loss_fn (_type_): the loss value of a batch
"""
super(LogSoftmaxWrapper, self).__init__()
self.loss_fn = loss_fn
self.criterion = paddle.nn.KLDivLoss(reduction="sum")
def forward(self, outputs, targets, length=None):
targets = F.one_hot(targets, outputs.shape[1])
try:
predictions = self.loss_fn(outputs, targets)
except TypeError:
predictions = self.loss_fn(outputs)
predictions = F.log_softmax(predictions, axis=1)
loss = self.criterion(predictions, targets) / targets.sum()
return loss
class NCELoss(nn.Layer):
"""Noise Contrastive Estimation loss funtion
Noise Contrastive Estimation (NCE) is an approximation method that is used to
work around the huge computational cost of large softmax layer.
The basic idea is to convert the prediction problem into classification problem
at training stage. It has been proved that these two criterions converges to
the same minimal point as long as noise distribution is close enough to real one.
NCE bridges the gap between generative models and discriminative models,
rather than simply speedup the softmax layer.
With NCE, you can turn almost anything into posterior with less effort (I think).
Refs:
NCEhttp://www.cs.helsinki.fi/u/ahyvarin/papers/Gutmann10AISTATS.pdf
Thanks: https://github.com/mingen-pan/easy-to-use-NCE-RNN-for-Pytorch/blob/master/nce.py
Examples:
Q = Q_from_tokens(output_dim)
NCELoss(Q)
"""
def __init__(self, Q, noise_ratio=100, Z_offset=9.5):
"""Noise Contrastive Estimation loss funtion
Args:
Q (tensor): prior model, uniform or guassian
noise_ratio (int, optional): noise sampling times. Defaults to 100.
Z_offset (float, optional): scale of post processing the score. Defaults to 9.5.
"""
super(NCELoss, self).__init__()
assert type(noise_ratio) is int
self.Q = paddle.to_tensor(Q, stop_gradient=False)
self.N = self.Q.shape[0]
self.K = noise_ratio
self.Z_offset = Z_offset
def forward(self, output, target):
"""Forward inference
Args:
output (tensor): the model output, which is the input of loss function
"""
output = paddle.reshape(output, [-1, self.N])
B = output.shape[0]
noise_idx = self.get_noise(B)
idx = self.get_combined_idx(target, noise_idx)
P_target, P_noise = self.get_prob(idx, output, sep_target=True)
Q_target, Q_noise = self.get_Q(idx)
loss = self.nce_loss(P_target, P_noise, Q_noise, Q_target)
return loss.mean()
def get_Q(self, idx, sep_target=True):
"""Get prior model of batchsize data
"""
idx_size = idx.size
prob_model = paddle.to_tensor(
self.Q.numpy()[paddle.reshape(idx, [-1]).numpy()])
prob_model = paddle.reshape(prob_model, [idx.shape[0], idx.shape[1]])
if sep_target:
return prob_model[:, 0], prob_model[:, 1:]
else:
return prob_model
def get_prob(self, idx, scores, sep_target=True):
"""Post processing the score of post model(output of nn) of batchsize data
"""
scores = self.get_scores(idx, scores)
scale = paddle.to_tensor([self.Z_offset], dtype='float64')
scores = paddle.add(scores, -scale)
prob = paddle.exp(scores)
if sep_target:
return prob[:, 0], prob[:, 1:]
else:
return prob
def get_scores(self, idx, scores):
"""Get the score of post model(output of nn) of batchsize data
"""
B, N = scores.shape
K = idx.shape[1]
idx_increment = paddle.to_tensor(
N * paddle.reshape(paddle.arange(B), [B, 1]) * paddle.ones([1, K]),
dtype="int64",
stop_gradient=False)
new_idx = idx_increment + idx
new_scores = paddle.index_select(
paddle.reshape(scores, [-1]), paddle.reshape(new_idx, [-1]))
return paddle.reshape(new_scores, [B, K])
def get_noise(self, batch_size, uniform=True):
"""Select noise sample
"""
if uniform:
noise = np.random.randint(self.N, size=self.K * batch_size)
else:
noise = np.random.choice(
self.N, self.K * batch_size, replace=True, p=self.Q.data)
noise = paddle.to_tensor(noise, dtype='int64', stop_gradient=False)
noise_idx = paddle.reshape(noise, [batch_size, self.K])
return noise_idx
def get_combined_idx(self, target_idx, noise_idx):
"""Combined target and noise
"""
target_idx = paddle.reshape(target_idx, [-1, 1])
return paddle.concat((target_idx, noise_idx), 1)
def nce_loss(self, prob_model, prob_noise_in_model, prob_noise,
prob_target_in_noise):
"""Combined the loss of target and noise
"""
def safe_log(tensor):
"""Safe log
"""
EPSILON = 1e-10
return paddle.log(EPSILON + tensor)
model_loss = safe_log(prob_model /
(prob_model + self.K * prob_target_in_noise))
model_loss = paddle.reshape(model_loss, [-1])
noise_loss = paddle.sum(
safe_log((self.K * prob_noise) /
(prob_noise_in_model + self.K * prob_noise)), -1)
noise_loss = paddle.reshape(noise_loss, [-1])
loss = -(model_loss + noise_loss)
return loss
class FocalLoss(nn.Layer):
"""This criterion is a implemenation of Focal Loss, which is proposed in
Focal Loss for Dense Object Detection.
Loss(x, class) = - \alpha (1-softmax(x)[class])^gamma \log(softmax(x)[class])
The losses are averaged across observations for each minibatch.
Args:
alpha(1D Tensor, Variable) : the scalar factor for this criterion
gamma(float, double) : gamma > 0; reduces the relative loss for well-classified examples (p > .5),
putting more focus on hard, misclassified examples
size_average(bool): By default, the losses are averaged over observations for each minibatch.
However, if the field size_average is set to False, the losses are
instead summed for each minibatch.
"""
def __init__(self, alpha=1, gamma=0, size_average=True, ignore_index=-100):
super(FocalLoss, self).__init__()
self.alpha = alpha
self.gamma = gamma
self.size_average = size_average
self.ce = nn.CrossEntropyLoss(
ignore_index=ignore_index, reduction="none")
def forward(self, outputs, targets):
"""Forword inference.
Args:
outputs: input tensor
target: target label tensor
"""
ce_loss = self.ce(outputs, targets)
pt = paddle.exp(-ce_loss)
focal_loss = self.alpha * (1 - pt)**self.gamma * ce_loss
if self.size_average:
return focal_loss.mean()
else:
return focal_loss.sum()
class GE2ELoss(nn.Layer):
"""Generalized end-to-end loss which defined in the paper "GENERALIZED END-TO-END LOSS FOR SPEAKER VERIFICATION"
"""
def __init__(self, init_w=10.0, init_b=-5.0, loss_method="softmax"):
super(GE2ELoss, self).__init__()
self.loss_method = loss_method.lower()
self.w = self.create_parameter(
[1], default_initializer=I.Constant(init_w))
self.b = self.create_parameter(
[1], default_initializer=I.Constant(init_b))
assert self.loss_method in ["softmax", "contrast"]
def get_cossim(self, embeddings_list, centroids):
"""Compute cosine similarity for each speaker
"""
cossims = []
for s_idx, embeddings in enumerate(embeddings_list):
cossim = F.linear(embeddings, centroids.t())
e_num = len(embeddings)
if embeddings.ndim > 1 and e_num > 1:
expand_centroids = paddle.expand(
centroids[s_idx], shape=[e_num, embeddings.shape[1]])
new_centroids = (expand_centroids * e_num - embeddings) / (
e_num - 1)
sims = F.cosine_similarity(embeddings, new_centroids)
cossim[:, s_idx] = sims
cossims.append(self.w * cossim + self.b)
return cossims
def cal_softmax_loss(self, cossims):
"""Calculate softmax loss
"""
loss = 0.0
n = 0
for s_idx, cossim in enumerate(cossims):
loss += -F.log_softmax(cossim, axis=1)[:, s_idx].sum()
n += cossim.shape[0]
return loss / n
def cal_contrast_loss(self, cossims):
"""Calculate contrast loss
"""
loss = 0.0
n = 0
for s_idx, cossim in enumerate(cossims):
cossim = F.sigmoid(cossim)
col_loss = 1. - cossim[:, s_idx]
if len(cossims) > 1:
if s_idx == 0:
excl_centroids_sigmoids = cossim[:, s_idx + 1:]
elif s_idx == (len(cossims) - 1):
excl_centroids_sigmoids = cossim[:, :s_idx]
else:
excl_centroids_sigmoids = paddle.concat(
(cossim[:, :s_idx], cossim[:, s_idx + 1:]), axis=1)
col_loss += paddle.max(excl_centroids_sigmoids, axis=1)[0]
loss += col_loss.sum()
n += cossim.shape[0]
return loss / n
def forward(self, output, target):
"""Forward inference
Args:
output: input tensor
target: target label tensor
"""
spkers = paddle.unique(target)
embeddings_list = []
for spkid in spkers:
index = (target == spkid).nonzero().reshape([-1])
embeddings_list.append(output[index])
# cal centroid
centroids = []
for embeddings in embeddings_list:
if (embeddings.ndim > 1):
spker_centroid = paddle.mean(embeddings, axis=0)
else:
spker_centroid = embeddings
centroids.append(spker_centroid.clone() / paddle.norm(
spker_centroid, axis=0, keepdim=True))
centroids = paddle.stack(centroids)
# cal cosine similarity
cossims = self.get_cossim(embeddings_list, centroids)
# cal loss
if self.loss_method == "softmax":
loss = self.cal_softmax_loss(cossims)
else:
loss = self.cal_contrast_loss(cossims)
return loss
if __name__ == "__main__":
import numpy as np
from paddlespeech.vector.utils.vector_utils import Q_from_tokens
paddle.set_device("cpu")
input_data = paddle.uniform([32, 100], dtype="float64")
label_data = np.random.randint(0, 4, size=(32)).astype(np.int64)
input = paddle.to_tensor(input_data)
label = paddle.to_tensor(label_data)
loss1 = GE2ELoss(loss_method="softmax")
loss = loss1.forward(input, label)
print("GE2ELoss softmax-loss: %.5f" % (loss[0]))
loss2 = GE2ELoss(loss_method="contrast")
loss = loss2.forward(input, label)
print("GE2ELoss contrast-loss: %.5f" % (loss[0]))
loss3 = FocalLoss()
loss = loss3.forward(input, label)
print("FocalLoss loss: %.5f" % (loss))
Q = Q_from_tokens(100)
loss4 = NCELoss(Q)
loss = loss4.forward(input, label)
print("NCELoss loss: %.5f" % (loss))