IR7;ve seen research going on for using to encode of (in a video, for purposes like classification, etc). I wonder, for LSTM in general, is there a limit on how long each example should be? I have a dataset where each point consists of 3 (each 1-2 seconds apart) and was trying to see if LSTM can be used to encode them.



Source link
thanks you RSS link
( https://www.reddit.com/r//comments/9p9xh4/d_lstm_for_sequence_of_images/)

LEAVE A REPLY

Please enter your comment!
Please enter your name here