Fig. 8From: Fully convolutional networks for structural health monitoring through multivariate time series classificationExample of time evolutions of z displacements for stories 1, 4, 8 with SNR\(=15\) dB (from a–f) and SNR\(=10\) dB (from g–l), undamaged state. Low-noise case: \(f_{1,2}^{ax}=\left( 32.8, 28.2 \right) \), \(\gamma _{1,2}^{ax}=\left( 1.38,1.38\right) \). High-noise case: \(f_{1,2}^{ax}=\left( 15.5, 22.0 \right) \), \(\gamma _{1,2}^{ax}=\left( 1.133,-1.140\right) \). Orange lines represent \({\varvec{u}}\), whereas black lines stand for \({\varvec{r}}\), according to Eq. (2). On the right side, a closer view for each left side plot is reportedBack to article page