No, no, no. The point is feeding the network slightly different images as they're captured by a small field of vision. The real perceived object is not static.

No, no, no. The point is feeding the network slightly different images as they're captured by a small field of vision. The real perceived object is not static. You're going to need a temporal, recurrent type of model. Also: a bigger boat.