PYTORCH支持的torch.nn.CTCLoss和torch_baidu_ctc支持的CTCLoss有区别吗？

Question

PYTORCH支持的torch.nn.CTCLoss和torch_baidu_ctc支持的CTCLoss有区别吗？

我想，当我比较教程代码时，我没有注意到任何区别。

有谁知道真相吗？

教程代码位于下面。

import torch
from torch_baidu_ctc import ctc_loss, CTCLoss

# Activations. Shape T x N x D.
# T -> max number of frames/timesteps
# N -> minibatch size
# D -> number of output labels (including the CTC blank)
x = torch.rand(10, 3, 6)
# Target labels
y = torch.tensor([
# 1st sample
1, 1, 2, 5, 2,
# 2nd
1, 5, 2,
# 3rd
4, 4, 2, 3,
],
dtype=torch.int,
)
# Activations lengths
xs = torch.tensor([10, 6, 9], dtype=torch.int)
# Target lengths
ys = torch.tensor([5, 3, 4], dtype=torch.int)

# By default, the costs (negative log-likelihood) of all samples are 
summed.
# This is equivalent to:
#   ctc_loss(x, y, xs, ys, average_frames=False, reduction="sum")
loss1 = ctc_loss(x, y, xs, ys)

# You can also average the cost of each sample among the number of 
frames.
# The averaged costs are then summed.
loss2 = ctc_loss(x, y, xs, ys, average_frames=True)

# Instead of summing the costs of each sample, you can perform
# other `reductions`: "none", "sum", or "mean"
#
# Return an array with the loss of each individual sample
losses = ctc_loss(x, y, xs, ys, reduction="none")
#
# Compute the mean of the individual losses
loss3 = ctc_loss(x, y, xs, ys, reduction="mean")
#
# First, normalize loss by number of frames, later average losses
loss4 = ctc_loss(x, y, xs, ys, average_frames=True, reduction="mean")


# Finally, there's also a nn.Module to use this loss.
ctc = CTCLoss(average_frames=True, reduction="mean", blank=0)
loss4_2 = ctc(x, y, xs, ys)

# Note: the `blank` option is also available for `ctc_loss`.
# By default it is 0.

torch.nn.CTCLoss

T = 50      # Input sequence length
C = 20      # Number of classes (excluding blank)
N = 16      # Batch size
S = 30      # Target sequence length of longest target in batch
S_min = 10  # Minimum target length, for demonstration purposes

# Initialize random batch of input vectors, for *size = (T,N,C)
input = torch.randn(T, N, C).log_softmax(2).detach().requires_grad_()

# Initialize random batch of targets (0 = blank, 1:C+1 = classes)
target = torch.randint(low=1, high=C+1, size=(N, S), dtype=torch.long)

input_lengths = torch.full(size=(N,), fill_value=T, dtype=torch.long)
target_lengths = torch.randint(low=S_min, high=S, size=(N,), 
dtype=torch.long)
ctc_loss = nn.CTCLoss()
loss = ctc_loss(input, target, input_lengths, target_lengths)
loss.backward()

我是韩国人。英语不是我的第一语言。所以我的英语不太好。如果有任何东西没有很好地交付，请留下评论。我会尽快修改句子的。

Answer 1

自1.0 version以来，CTC只损失了PyTorch的一部分，而且它是一种更好的方法，因为它本身就是PyTorch的一部分。如果您使用的是PyTorch 1.0或更高版本，请使用torch.nn.CTCLoss。

warp-ctc似乎没有得到维护，最后一次提交更改核心代码是在2017年。后来，他们只修复了TensorFlow (一个已经过时的版本)的绑定。

Answer 2

注意Pytorch CTC loss将log softmax概率作为输入，而百度的CTC loss没有提到这一点。