Fig. 1
From: Fundamentals for predicting transcriptional regulations from DNA sequence patterns
Conceptual representation of convolution operation. A Toy example of convolution operation using a 7 × 7 input image (original image) and 3×3 kernel. This image and kernel have two colors: 1 is red, and 0 is white in the input image, and 1 is green and 0 is white in the kernel. In the feature map, the output value for the location is highlighted in red. In this toy example, the bias parameter is not shown. B, C There are two examples of convolution using a kernel with one in the upper left and 0 in others (B) and a “C”-like kernel (C)
