What did I do this past week?

Past week I finished up voting assignment and did some homework for my marketing class. I also went to a taekwondo competition this Saturday at Texas A&M. It was my first competition and I competed as a white belt. Honestly I got beaten up pretty hard. My opponent was a lot heavier, faster and stronger than I am and he knew how to fight. He got a lot of head kicks on me and I was too tense to block his kicks but it was a good experience and had good time spending time with the team overall.

What’s in my way?

I’m still a little devastated I lost so hard so I need get my mindset ready for next week.

What will I do next week?

I will start my computer vision assignment. Also there is NBA All star events I missed last week so I will catch up on that stuff. If I get more free time I might do some extra competitive programming problems.

What’s my experience of the class?

Last week we learned about generic types and different types of iterators. Generic types are super useful and I can definitely see myself using them when I write C++ code. The thing about iterators I still can’t understand is that we give them names as input, output, forward and so on but are these official names and how are they useful other than our naming purposes of our variables. If I make my own class and make it forward iterable do I need it implement forward iterable interface or something?

Pick of the week

Last week I used Shazam to find a song I heard and it got me thinking how does a software identify a song so fast when there are literally millions of songs out there. After ding some research (googling for 5 minutes) I found some good explanations. Shortly Shazam generates some kind of ‘fingerprint’ of songs and these fingerprints are some kind of simple numeric values. How does Shazam creates a simple numeric value to identify a song is the interesting part. It is done using a spectrogram, a graph that plots frequency vs. amplitude vs. time. In practice Shazam creates 3 data points for per second per song. So by listening just a few seconds of a song they create 6 to 9 data points and match these points to the ones in the database. A more detailed explanation could be found here.