Good evening, ive beens struggling training a dataset made of 80x60 matrix decoded with 32_FC1.
I have 2 datasets files one of positive and one of negative and i want to train a CNN with keras and Adam with these 2 datasets but its impossible for me to do the “input” of that 2 datasets to create that binary image classificator. I dont know how to code that input because all the depth images decoded in CV_32FC1 are in the same file separated with one tab.
Thanks in advance. If u need further information ask me please and sorry for my bad english.
I have one dataset with sewer depth images and another dataset with no sewer images, all images are depth type and decoded in 32FC1, and now with these two dataset I want to train a CNN that could tell me if a image has a sewer in it or not.
Each dataset its like a 8000x60 32CF1 matrix with 80x60 images in it. So there are like 100 depth images inside each dataset.
Sorry but thats why i posted this i dont know how to start i have not found any post or website that have done this before. They all start with a labeled dataset of png photos and now 2 different datasets with decoded images inside.
If u could tell me at least one link or post so i could start i would be apreciated.
Im just asking if opencv folks have ever work with Open CV type C3FC1 with Keras, U dont need to be that rude im just asking for help if this is not the site im sorry just tell that.
no its not rude (no worries)
however, we can better help you, if you can be more specific
(and, imo, its too early now, and you’re on the wrong bus here)
Hahaha this is my final work to finish my degree but ive been working on this too many months i will change my work next year. You are a good person mate