This week we started to integrate the lua model onto the nano board, but as most things have gone for us this year we ran into some issues, one of which being the fact that the model has too many layers for the nano to handle. This is a major Read more…
This week there was some real concerning realization that our group had to face, including the fact that the originally planned CNN model struggles to learn. One thing I attempted was changing the size of the input packet to be of size 1 instead of a stack of 5 images, Read more…
This week there was a realization that the CNN model was not working as intended and was really not learning during the training. I trained on a smaller batch size and was getting the same weird results, I am about to start training it with a larger batch size and Read more…
This week there were plenty of hurdles for the deblurnet CNN and I’m sure there will be more to come. The training code was finally finished and the data was properly split into frames of 128×128 and all that was left was to deploy the ec2 instance and begin training. Read more…
This week we decided that it was not worth it to wait for the original SD card that we ordered, so we put a request in for another SD card of smaller memory from Amazon, a much more reliable source than Wish.com, and the SD card came the next day. Read more…
This week we were finally able to take the whole week to work on implementation. So I am tasked with creating and training the DeBlurNet CNN that will take in a packet of frames and then produce a singular deblurred frame. We decided to go with pytorch instead of tensor Read more…
This week we were scrambling a little it because of how quick the turnaround was from proposals to an actual design. About halfway through the week we met as a team and got some ideas flowing for our block diagram and some of the smaller implementation details. We were essentially Read more…
This week was sort of a bigger jump for us because we really started to get more technical and start to think about implementation details and really hammer down some of the bigger pictures. To start we began by making an overall design structure and block diagram which we started Read more…
This week we were able to get more insight on what our CNN design should look like, feedback from our professors and TA. From this new direction we looked more into encoder/decoder style and got a better understanding of what we should do. We were also able to look more Read more…
So this week consisted of a lot of research for me, I started off by looking at most of the relevant resources from the original paper we are building our CNN from, project website found here. The first thing I looked into was building a regular CNN that would take Read more…