In order to understand a Deep Learning Network’s architecture it’s better to visualize it. Probably you are a TensorFlow user and you have Tensorboard to visualize your trainning process, haven’t use it much but I probably will do once I set it up for PyTorch. But just check out Nvidia Digit’s beauty when visualizing inferences: Anyways, for networks themselves I loved Nvidia’s Digits due to the visualizations in inference and specially in the network (As seen in the header in the left part) and was looking into similar tools for other environments that are not Digit’s attached.

If you didn’t attend this year GTC (GPU Technology Conference) in Washington DC worry not, in this post I’ll try to summarize it since I got the chance to go. Quoting it from the page itself: “Presented by NVIDIA, the GPU Technology Conference is the world’s most important event for GPU developers. Experience five days of hands-on AI training, industry insights, and direct access to NVIDIA experts-all in one place.” Let me tell you that they delivered!

(Most of the images are taken from Udacity’s SDCND page, all credit to them) I think I am lucky to say that I am a mentor for Udacity’s first ever Self Driving Car Nanodegree. Some quick background on me, I am a graduate student at WPI (Worcester Polytechnic Institute) here in Massachusetts pursuing a master in robotics engineering. Well some might argue that it’s not the first course, since there is Artificial Intelligence for Robotics, but to be honest this Nanodegree dives deep into the concepts of Machine Learning, Computer Vision, Localization, Control, etc.

I am part of the WPI Team that is competing in the Space Robotics Challenge by Nasa (Credits to them for the Valkyrie image on your left) and in this small post I’ll teach you to: Automate your testing with Gitlab-CI Prepare your Runners (The machines that run your tests) Get notifications using a SlackBot to know when your build is broken And you’ll say why Gitlab and not Github? Everybody loves Github right?

Working with a humanoid robot was an awesome new experience for me. Some background, due to a Fulbright scholarship I am able to do a master in robotics at the Worcester Polytechnic Institute (WPI) here in Massachusetts. I’ve started the semester and everything here in the labs is amazing, there is an Atlas Robot from Boston Dynamics (picture on the left), a PR2, a Baxter, Kukas, Fanucs, ABBs, everything that a robotics would love to experiment with.

So this year is almost done and I thought I would summarize it in some lines. This 2014 was awesome, it really was. Even though I didn’t do most of the things I wanted, I would say I did the best ones. First things first, I finished my career… kind of… things over here are not always what one would like, bureaucracy made my life quite complicated but in the end it doesn’t even matter.