-
Notifications
You must be signed in to change notification settings - Fork 58
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
train stage #11
Comments
肯定不可以这样操作,如果loss是估计回声和真实回声的欧式距离,那么反向的时候梯度肯定要对数学建模部分的卡尔曼滤波进行反向,我的建议是计算数学方法的凯尔曼滤波器系数和网络输出的滤波器系数的距离。或者固定数学建模部分,使用最后的loss。 |
那作者怎么能训起来呢,请问你已经跑通了训练过程吗 |
我自己写过,可以进行训练。作者还是很牛逼的。 |
|
尴尬咯,我的实验结果是,基础的数学建模凯尔曼滤波 < 网络的凯尔曼滤波 < 状态因子优化的数学建模凯尔曼滤波 |
论文里一句话"After one forward pass" 估计回声hat_h和消除后的信号hat_S 可以得到, 然后用真实回声和估计回声做loss。 这里面的 one forward pass 难道是一个时刻t? 每一帧算一次loss? 然后更新梯度? |
我查到了作者之前关于这篇论文的报告 https://www.livevideostack.cn/video/ |
你好,请问一下你成功复现出了模型效果吗 |
您好,可以参考一下您的训练代码吗?我自己写的经常出现nan的情况,感谢! |
|
请教下,状态因子优化的数学建模卡尔曼滤波是怎么做的?有相关论文没 |
有人复现出训练过程么,方便请教么 |
大神你好,你是怎么训练的呢?教教我 |
我写了好几种训练方法,已经可以训练了,嘻嘻嘻嘻 |
|
|
|
Can the training stage directly use this forward code to train the model? What is there to pay attention to? Thanks!
The text was updated successfully, but these errors were encountered: