Category Archives: Uncategorized

After the 50th Post, some vision on the Grand Janitor's Blog

I tried to gather some topic ideas on this blog.   From LinkedIn, I hear quite a bit of feedbacks on discussion about decoding algorithms.    I have thought about this a bit.   Here are some sample topics which I will discuss,

  • How to setup Recognizer X on a Platform Y? 
  • What are the interesting properties of Program A in speech recognition package B? 
  • What are the best practice when you try to use a certain model training package? 
  • Tuning results in a domain. 
  • News: certain package has a new version coming out! What are something special? 
  • Thoughts in speech recognition problem.
  • Thoughts in programming problem.
  • Algorithms/Data structures for decoding/parameter estimation : like you suggested in this post. 
  • What is an interesting development direction for a speech recognition package?
  • Others: softwares/algorithm in some related fields such as NLP, MT. 

Of course, I will keep on chatting about some on-going components of Sphinx which I feel interested,

  1. Sphinx 4 maintenance and refactoring
  2. PocketSphinx's maintenance
  3. An HTKbook-like documentation : i.e. Hieroglyphs. 
  4. Regression tests on all tools in SphinxTrain.
  5. In general, modernization of Sphinx software, such as using WFST-based approach.
Unlike academic journals, I hope this is more an informal hub of ASR information.   I don't know if I can keep on. but at least that's what I am going to try. 
In any case, this blog will occupy much of my spare time next year.   Feel free to comments and give me suggestions.  And of course ....... Happy New Year!

Arthur

Random Thought: Cloud


When I was in College in Hong Kong, I love to stare at the blue sky and just watching pieces of cloud floating from my left to right. There was much open space in the University. My favorite thing to do is to skip classes and watch some clouds.

To many of my friends, that is a ridiculous habit. Though most of them see them as part of my little eccentricities in my little unsung college career.

In another words, I have done worse. πŸ™‚ So they are not truly surprised and I am not that disappointed by their misunderstanding of clouds.

My true disappointment comes when I tried to share this interesting hobby with a mathematically-oriented friend. This guy is genuinely smart. In terms of Math, I think he is about 5 years ahead of me. So I thought he would understand.

So I told him my true intention of watching cloud - I would like to predict weather based on observing the cloud. That, to me, is a totally reasonable application of Mathematics. This is his response,

"You read "Wind and Cloud" too much.".

("Wind and Cloud" is a popular martial art comic book in Hong Kong. It's about a two martial experts, "Wind" and "Cloud" and their adventure in China.)

Many people asked me why I chose to live in US instead of Hong Kong, or even Bigger China. This story is probably an example of why.

In Hong Kong (or probably the bigger China), it is a difficult thing for students to imagine that advanced mathematics could have anything to do with complex subjects such as metereology at all. Also, there is a big gap between the expert knowledge of a certain field and the general public. So even if you have a technical background and you are smart enough to learn, you could still be ignorant on branches of other fields.

Of course, an even deeper problem is that imagination and creativity is not an emphasis in technical subjects such as Science and Mathematics. In the secondary school curriculum, they were usually not taught to inspire students to discover Mathematics themeslves. This explains the behavior of my smart friend.

There are social consequences of this, students grow up like this will probably unable to appreciate interesting thought from the youngs. That is to say scientific and technical workers are not truly appreciated. This compounds with the general money-loving attitude in Hong Kong. You will not surprised that Science and Technology is tough to develop there.

We cannot say the States' education is perfect, there are tons of holes and problems in it as well. But perhaps because Americans are always more adventurous in nature. They always see possibilities. That's why if you asked a smart student in U.S. the same question, you would probably got an account of General Circulation Model, how the basic equations is written. How Stoke-Navier equation can be used in this problem. (If we digress, then we would chat about how Stoke-Navier equation could be one of the 7 Millenium problems.)

I don't resent my friend's comment. What I see was that a smart person like him was wasted in the system. How many more of these situations happened in the past? I have no idea. What I know is that this is the true impedance of generating good scientific and technical workers.

Statistically Insignificant Me

Slightly related my last post. It relates to an interesting issue of whether we should share the bookshelf in the first place.

Why is it an issue? Well, privacy. Suppose someone is malicious and try to figure you out. The best way is to try to gather all information about you and work against you.

Another concern of mine is rather interesting and absolutely speculative, what if information I read will affect my thought and what if people could reconstruct it just from the information I read? That will open up a lot of interesting application. e.g. We might be able to predict what a person will do better.

Just like in other time series problem such as speech recognition and quantitative analysis. Human life could simply be defined by a series of time events. Some (forget the quote) believes that one human life could be stored in hard-disk and some starts to collect human life and see whether it could be model.

Information of what you read could tell a lot of who you are. Do you read Arthur C. Clarke? Do you read Jane Austen? Do you read Stephen King? Do you read Lora Roberts? From that information, one could build a machine learner to reverse map to who you are and how you make decision. We might just call this a kind of personality modeling.

It seems to me these are entirely possible from the standpoint of what we know. Yet, I still decide to share my bookshelf? Why?

Well, this was crystal-clear moment for me (and perhaps for you as well) which helps me to make a decision: Very simple, *I* am statistically in-significant.

If you happen to come to this web page, the only reason you come is because you are connected to me. How likely will that happened?

I know about 150 persons in my life. The world has about 6 billion. So that simply means the chance of me being discovered is around 1.5 x 10^-8. It is already pretty low.

Now, when other people know me and recommend me to someone else. Then this probability will be boosted up because 1) my PageRank will increase, 2) people follow my link deep enough will eventually discovered my bookshelves.

Yet, if I try to stay low-profile, (say not try to do SEO, not recommend any friends to go to my page) then it is reasonable to expect the factor mentioned is smaller than 1.

Further, 1.5 x 10^-8 is an upper bound as an estimate because
1, Not all my friends are interested in me (discounting factor : 0.6, a conservative one, the actual number is probably higher but I just don't want to face it. πŸ˜‰ )
2, My friends who are interested in me might not follow my links (discounting factor: 0.01)

So we are talking about an event with probability as low as 10^-9 or 10^-10 here. That seems to me close to cheap cryptographic algorithm.

But notice here, my security is not come from hiding or cryptography. My security merely come from my statistical insignificance. In English, I am very open but no one cares. And I am still a happy treebear. πŸ˜‰

That's why you see my bookshelf. Long story for a simple decision. If you happen to read this, I hope you enjoy it.

-a

Visual Bookshelves

I love to read and like to write reviews for every books I read. None of them will change the world but it still loves to do it. That's why by definition - I'm a bookworm. Not even feel shy about it. πŸ˜‰

I go quite far: try to record every books I read on a blog and start to put them in a blog called "ContentGeek". Luckily, I haven't gone very far. Because once I discovered Visual Bookshelves, there is no need for me to do it all.

Visual Bookshelves allow users to look up a book from Amazon, add comments and stored it in a database. It also shows the book cover of the books. What else could I want more?

So anyway, this is the link of my visual bookshelves:

http://www.cs.cmu.edu/~archan/personal/bookshelf.html

Enjoy.

-a

David's plan on Sphinx 3.7

http://lima.lti.cs.cmu.edu/mediawiki/index.php/Sphinx3

A great read, it touches the heart of implementation issues of all sphinxen. And its criticism on my implementation a right straight to the point.

I felt very relieved when the current maintainer attack what I did in the past. (Some features I did were rather stupid.) This shows that Sphinx is still alive and will still be alive.

-a

Life in Scanscout

Hi Guys,
Scanscout (www.scanscout.com) is a rather interesting company. . If you look at this blog, you probably know that I have been there for a while.

My direct supervisor doesn't like to give away too much. I think he has a point (as he is a *v* smart guy"). This contradicts to my philosophy of information sharing. So alright, as a compromise, here are couple of things I could share. (Of course, my estimate of the probably of anyone looking at this blog is about 1/10^9, so I guess it doesn't matter that much......)

1, We have a massage chair and it is awesome.
2, We have a foozball table and have a tournament every Friday. Beware, there are several good players. (I always get the lowest score.)
3, It is on the fore-front of video advertising. I am glad that I've joined. πŸ™‚

Arthur Chan

mosedecoder

Mosedecoder: http://www.statmt.org/moses/

Ah. This is not exactly news. It has been around since 2006 John Hopkins workshop.

mosedecoder is probably the first open source statistical machine translation implementation in the world. For quite a while, only the IBM models training portion of the code could be found in GIZA++. So for people who is interested in SMT, they will probably turn to Pharaoh, a close source implementation available in the web.

I could have some fun. πŸ˜‰

-a

Third Draft of Hieroglyphs

Hi all,

It has been a while I worked on the Hieroglyphs (the fancy name I made for sphinx documentation). This is perhaps the only things I haven't wrapped up in CMU. Therefore I decided to release a draft. You can find it

at

http://www-2.cs.cmu.edu/~archan/documentation/sphinxDocDraft3.pdf

It still looks pretty messy but it starts to look like a book now.

Several chapters and sections were trimmed in this draft. You will still see a lot of ?. Those are signals of not enough proof-reading. Forgive me, when I have more time, I will try to fix some of them in near future.

Grand Janitor

Left CMU

Hi Guys,
It was a sad decision. After a long soul-searching, I decided to leave CMU and join a startup company called Scanscout. I must be out of mind!!

Anyway, my new job require knowledge in speech recognition, information retrieval and video processing. These are all good fit for me. I could tell you I have a lot of fun!

Sphinx, in particular the trio, Sphinx 3.X, SphinxTrain and CMULMTKV3 are now maintained by David Huggins-Daines and Evandro Gouvea. I still keep a nominal maintainership but these two are the true heros in the story now.

However, feel free to chat with me on anything related language processing. I am more than happy to be there.

Regards,
Arthur Chan