Mapping speech timestamps to windows or segments

[jarashanth]

Thank you for your good work. It's really helpful for beginners like me. I have the following doubt.
for i in speech_timestamps:
start = int(min((i['start']+win_step/2)//win_step, NumWin))
end = int(min((i['end']+win_step/2)//win_step, NumWin))
speech_segments[start:end] = 1

Here, I do not understand how the mapping to start and end works. Isn't it supposed to be i[''start"]+win_len/2?