It may give them a permanent taste for what is abstruse and infinitely subtle — at the expense of anything that doesn't pass an exaggerated threshold of convolution.

这可能会让他们对深奥和无限微妙东西产生永久兴趣——以牺牲所有没有通过夸大卷积阈值东西为代价。

The weapon of choice for this work was a Convolutional Neural Network, and the dataset the algorithm was trained on couldn't be simpler: it was given a large database of 2D input image and 3D output geometry pairs.

这项工作选择算法是卷积神经网络，并且常用算法数据集已经被精简不能再简单了：它只有一个大2D输入图像数据库，搭配3D几何输出模型。

Powered by the massive data from ImageNet and the modern CPUs and GPUs to train such a humongous model, the convolutional neural network blossomed in a way that no one expected.

由ImageNet 提供巨大资群、并使用先进核心处理器及图型处理器来训练 这个庞然大物，卷积神经网络人意花结果了。

Machine learning kernels such as LSTMs, convolutions, and neuron layers are included as well as an expanded set of matrix math operations.

机器学习内核，比如 LSTMs、卷积码和神经元也包括内，还有扩展矩阵数学操作。

There is a huge amount of information that the convolutional neural network has to make sense of.

这一堆海量信息需要卷积神经网络来处理。

However, this is not your classical Convolutional Neural Network, because as we mentioned, the input is 2D and the output is 3D.

然而，这并不是传统卷积神经网络算法，因为正如我们曾提到，它输入是2D，输出是3D。

But not this time, because here, super effective descriptors are proposed automatically via a learning algorithm, a convolutional neural network and siamese networks.

但是这次不一样了，因为通过一种学习算法、卷积神经网络和 siamese 网络，而自动生成了超级有效描述符。

This is another great testament to the power of convolutional neural networks and as you can see, the rate of progress in AI and machine learning research is absolutely stunning.

这又是一个卷积神经网络坚实佐证，你可以看到人工智能和机器学习发展速度着实令人惊艳。

Since we have a neural network over here, it goes without saying that the training takes place on a large number of before-after image pairs, so that the network is able to produce these convolution kernels.

既然我们这里有了神经网络，这个训练理所当然地发生于大量前后相邻像对中，因此这个网络能够制成这些卷积核心程序。

This work uses a 3D convolutional neural network to accomplish this task.

研究中使用了三维卷积神经网络。

So the main innovation was saying, we're going to represent the state of the board with a neural net and convolutional neural nets are good at representing pixels and Go pieces on a board are kind of like pixels.

因此，主要创新于提出：我们将用神经网络来表示棋盘状态，而卷积神经网络擅长处理像素，围棋棋盘上棋子某种程度上类似于像素。

It's much more obvious that F convolved with G is the same thing as G convolved with F. Now, technically, the diagonal slices are not exactly the same shape.

F与G卷积与G与F卷积是相同， 这一点显而易见。不过，从技术上讲， 对角线切片形状并不完全相同。

Each one of those little 3D instances gets fed into a powerful pre-trained deep learning model, like a 3D convolutional neural network, CNN, or maybe a 3D vision transformer.

每一个微小3D实例都被输入到一个强大预训练深度学习模型中，比如3D卷积神经网络（CNN），或是3D视觉变换器。

They did this using something called a deep convolutional neural network and trained it with a dataset containing about 130,000 images.

他们使用一种叫做深度卷积神经网络东西来做到这一点，并用一个包含大约130,000张图像数据集来训练它。

At that time, the datasets were so small that those methods that used more hand engineering worked better than ConvNets.

当时，数据集非常小，那些使用更多手工设计方法比卷积神经网络效果更好。

In the application of simultaneous identification of gender and variety, CNN model has the highest accuracy of 94%, LDA model has the medium accuracy of 92.5%, and SVM model has the lowest accuracy of 89.5%.

性别与品种同时识别应用中，卷积神经网络（CNN）模型准确率最高，达到94%，线性判别分析（LDA）模型居中，准确率为92.5%，而支持向量机（SVM）模型准确率最低，为89.5%。

Convolutional Neural Networks (CNNs) excel in image-based tasks, while Recurrent Neural Networks (RNNs) and their variants, such as Long Short-Term Memory (LSTM) networks, excel in sequential data processing.

卷积神经网络 (CNN) 擅长基于图像任务，而递归神经网络 (RNN) 及其变体，例如长短期记忆 (LSTM) 网络，则擅长顺序数据处理。

This technique uses a convolutional neural network to accomplish something similar, but in the end, it doesn't give us an image, but a set of convolution kernels.

这项技术用了卷积神经网络以完成类似东西，但最终它并没有给出一副画面，而是一系列卷积核心程序。

But we could also make kernels that are good at finding lines, with edges on both sides.

但我们也可以制作擅长找到两边都有边缘线卷积核。

In image data, long-distance dependencies are typically modeled by the large receptive fields formed through deep stacks of convolutional operations.

图像数据中， 长距离依赖性通常由通过深度卷积运算堆栈形成大感受野建模。