These filters reduce the amount of light passing through the lens and means that the exposure time increases to compensate. With exposures of 20 or 30 seconds any people moving in the scene will not be in one place long enough to form an image and the resulting street scene will appear to be deserted.
For this type of result you need a fairly extreme neutral density filter – one with a density of 3.0 is good, giving a 10-stop reduction in light. This means that instead of shooting at 1/30 second, you can shoot with an exposure time of 32 seconds.
You will need to use the camera on a sturdy tripod to avoid movement during these long exposure times. However, the use of tripods is banned in the busy areas of some cities, so check before travelling.