S-Stem Scholar Monica Sanchez's Blog

I have continued with the lessons on Python this week and I also watched a few videos on Tensorflow. I learned that tensorflow can do other things besides image recognition. The program can also identify letters and other symbols. This aspect of Tensorflow means that it can be used to create a translator that translates words in a picture. I would like to explore this aspect of tensorflow as well, I am thinking that our next objective with this group will be to create a translator. Unfortunately, we are still having issues with the hardware. I have not worked with hardware as much, and I am more focused on the software, but by trying to fix the issues we are having, I am starting to learn a lot on the hardware side.

This week while my group member troubleshooted some issues with downloading the tensorflow, I focused more on learning more syntax with python. I used sololearn's python tutorials to learn more about python since I will need to use it to update the tensorflow models to fit the needs of our program. I really like how python is very similar to Java, so most of it also helps me refresh my memory on Java concepts. I have also watched some videos that explain how tensorflow works. Tensorflow is very complicated so these videos help simplify these complex ideas to me. For tensorflow, you use groups that classify images into categories, but I am wondering if it will end up being more simple for us since we just need to classify as correct and not correct. I am excited to test this theory when we get tensorflow running. First we will need to use sample models that test that tensorflow is working, then we can edit the code and the sample images to fit our needs. I believe that I will be abl...

Our group has been working on trying to download tensor flow onto a raspberry pi, but we keep getting different errors each time we try. We have found various models to use to train our program with when it gets running, but first we have to deal with trying to download tensor flow. It is a lot different than downloading something on windows. Usually you would just go to the website and hit the install button, but for this we have to issue commands telling the device what to do. After having many errors downloading tensor flow, we are now trying to update the system before we download tensor flow but we are also having issues updating it. We received an error saying resource unavailable, so we are currently trying to investigate how to get around that. Luckily there are many tutorials online that we are researching to help us move forward.