ID 原文 译文
17405 然而，研究表明脑电信息空间分辨率较低，这种缺陷可以综合分析多通道电极的脑电数据来弥补。 However, low spatial resolution isregarded as the deficiency of EEG signified from researches, which can fortunately be made up by syntheticanalysis of data from different channels.
17406 为了从多通道数据中高效地获取到与运动想象任务相关的辨识特征，该文提出一种针对多通道脑电信息的卷积神经网络(MC-CNN)解码方法， In order to efficiently obtain subspace features with discriminantcharacteristics from EEG channel information, a Multi-Channel Convolutional Neural Networks (MC-CNN)model is proposed for MI-EEG decoding.
17407 先对预先选取好的多通道数据预处理后送入2维卷积神经网络(CNN)进行时间-空间特征提取，然后利用自动编码(AE)器把这些特征映射为具有辨识度的特征子空间，最后指导识别网络进行分类识别。 Firstly input data is pre-processed form selected multi-channel signals,then the time-spatial features are extracted using a novel 2D Convolutional Neural Networks (CNN). Finally,these features are transformed to discriminant sub-space of information with Auto-Encoder (AE) to guide theidentification network.
17408 实验结果表明，该文所提多通道空间特征提取和构建方法在运动想象脑电任务识别性能和效率上都具有较大优势。 The experimental results show that the proposed multi-channel spatial featureextraction method has certain advantages in recognition performance and efficiency.
17409 随着深度学习技术的快速发展，利用深度神经网络模型伪造出的深度假脸(deepfake)视频越来越逼真，假脸视频造成的威胁也越来越大。 With the rapid development of deep learning technology, videos with changed faces generated bydeep neural networks (i.e., Deepfake videos) become more and more indistinguishable. As a result, the threatraised by Deepfake videos becomes greater and greater.
17410 文献中已出现一些基于卷积神经网络的换脸视频检测算法，他们在库内获得较好的检测效果，但跨库检测性能急剧下降，存在泛化能力不足的问题。 In literature, there are some convolutional neuralnetworks-based detection algorithms for fake face videos. Although those algorithms perform well when thetraining set and the testing set are from the same dataset, their performance could deteriorate dramatically incross-dataset scenario where the training and the testing sets are from different sources.
17411 该文从假脸篡改的机制出发，将视频换脸视为特殊的拼接篡改问题，利用流行的神经分割网络首先预测篡改区域，得到预测掩膜概率图，去噪并二值化， Motivated by the fabrication course of fake face videos, this article attempts to solve the problem of fake faces detection with the way of image splicing detection. A neural network borrowed from image segmentation is adopted for predicting the tampered face area from which a tampering mask is obtained through denoising and thresholding theprobability map.
17412 然后根据换脸主要发生在人脸区域的前提，提出一种计算人脸交并比的新方法，并进一步根据换脸处理的先验知识改进人脸交并比的计算，将其作为篡改检测的分类准则。 Using the prior knowledge of face tampering that the changing of face mainly happens in faceregion, a new way is proposed to determine the Face-Intersection over Union (Face-IoU) and to further improvethe ratio calculation method.
17413 所提出方法分别在3个不同的基础分割网络上实现，并在TIMIT, FaceForensics++, FFW数据库上进行了实验， The Face-Intersection over Union with Penalty (Face-IoUP) is used as theclassification criterion for deepfake video detection. The proposed method is impletmented using threebasic image segmentation neural networks separately and is tested them on datasets of TIMIT,FaceForensics++, Fake Face in the Wild(FFW).
17414 与文献中流行的同类方法相比，在保持库内检测的高准确率同时，跨库检测的平均错误率显著下降。 Compared with current methods in literature, the HTER(Half Total Error Rate) in cross-dataset test decreases significantly while the detection accuracy in intra-dataset test keeps high.