F3

Figure 3. Using multiple sequence alignment (MSA), we are able to optimally align input from multiple workers, then combine it into a single caption that is easy for the user to read.