Extended Data Fig. 2: The complexity of the CNN architecture controlled the tradeoff between speed and accuracy.
From: Segmentation of neurons from fluorescence calcium recordings beyond real time
We explored multiple potential CNN architectures to optimize performance. a-d, Various CNN architectures having depths of (a) two, (b) three, (c) four, or (d) five. For the three-depth architecture, we also tested different numbers of skip connections, ReLU (Rectified Linear Unit) instead of ELU (Exponential Linear Unit) as the activation function, and separable Conv2D instead of Conv2D in the encoding path. The dense five-depth model mimicked the model used in UNet2Ds9. The legend ‘0/ni + ni’ represents whether the skip connection was used (ni + ni) or not used (0 + ni). e, The F1 score and processing speed of SUNS using various CNN architectures when analyzing the ABO 275 μm dataset through ten-fold leave-one-out cross-validation. The right panel zooms in on the rectangular region in the left panel. Error bars are s.d. The legend (n1, n2, …, nk) describes architectures with k-depth and ni channels at the ith depth. We determined that the three-depth model, (4,8,16), using one skip connection at the shallowest layer, ELU, and full Conv2D (Fig. 1c), had a good trade-off between speed and accuracy; we used this architecture as the SUNS architecture throughout the paper. One important drawback of the ReLU activation function was its occasional (20% of the time) failure during training, compared to negligible failure levels for the ELU activation function.
