xbrainnet / CAFNet

IEEE Transactions on Affective Computing 2023
15 stars 4 forks source link

After processing all the subjects' EEG data, what is the characteristic matrix obtained? #7

Open logicvanlyf opened 4 months ago

logicvanlyf commented 4 months ago

Hello, after the preprocessing of the EEG data code you provided, I chose a s01.bdf in DEAP for processing. After processing, I got a one-dimensional array of shape (1600,). Excuse me, if you batch process the data of all subjects in DEAP, what is the final PSD feature matrix, such as its shape? If you can answer, I would be very grateful.

kaio-99 commented 4 months ago

值得注意的是,在DEAP数据集中,每个trial的长度应为60s,因此最终获得的特征维度应为40x20x32x5=40x20x160,其中20为时间窗口,32x5为通道数x频段数

logicvanlyf commented 4 months ago

你好,论文中提到取eeg数据的最后30秒用于后续特征提取,然后选取的是3秒非重叠窗口提取PSD,所以是10个窗口。那为什么是20个窗口呢?

---- 回复的原邮件 ---- | 发件人 | @.> | | 日期 | 2024年05月15日 15:10 | | 收件人 | @.> | | 抄送至 | @.>@.> | | 主题 | Re: [xbrainnet/CAFNet] After processing all the subjects' EEG data, what is the characteristic matrix obtained? (Issue #7) |

值得注意的是,在DEAP数据集中,每个trial的长度应为60s,因此最终获得的特征维度应为40x20x32x5=40x20x160,其中20为时间窗口,32x5为通道数x频段数

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

logicvanlyf commented 4 months ago

不好意思,我刚刚理解错意思了。处理mahnob数据集是每次使用trial的最后30秒,那使用DEAP数据集中的eeg数据时,是使用整个序列吗?

---- 回复的原邮件 ---- | 发件人 | @.> | | 日期 | 2024年05月15日 15:10 | | 收件人 | @.> | | 抄送至 | @.>@.> | | 主题 | Re: [xbrainnet/CAFNet] After processing all the subjects' EEG data, what is the characteristic matrix obtained? (Issue #7) |

值得注意的是,在DEAP数据集中,每个trial的长度应为60s,因此最终获得的特征维度应为40x20x32x5=40x20x160,其中20为时间窗口,32x5为通道数x频段数

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

logicvanlyf commented 4 months ago

您好,有时间能回答一下github上的新问题吗,不胜感激

---- 回复的原邮件 ---- | 发件人 | @.> | | 日期 | 2024年05月15日 15:10 | | 收件人 | @.> | | 抄送至 | @.>@.> | | 主题 | Re: [xbrainnet/CAFNet] After processing all the subjects' EEG data, what is the characteristic matrix obtained? (Issue #7) |

值得注意的是,在DEAP数据集中,每个trial的长度应为60s,因此最终获得的特征维度应为40x20x32x5=40x20x160,其中20为时间窗口,32x5为通道数x频段数

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>