Hey everyone.  As the holiday season approaches, I am finding time to gather some MATLAB code that I did for research projects that were completed.  While it is already mentioned on my project’s page, I thought I point to an update that code for “A New Distribution Metric for Image Segmentation” is now available.  You can also find it on the Mathworks MATLAB file exchange.  At any rate, here’s a quick link to download the code for those that are interested.

Download Link:  New_Distribution_Metric_Active_Contour_Segmentation.zip

As always, if you have any questions or comments, feel free to ask!

Day 3:

After giving our lecture on computer vision, Shawn and I began the actual site seeing.  At this point of the journey, we learned early on that the sun rose early about 5:30 AM, and that unlike back at home, we couldn’t do much touring around after 6:00 PM.  Of course, being graduate students, getting up early was no problem!  At any rate, we woke up at the crack of dawn and jumped into our rental car.  We headed up north to Rosh HaNikra, which is labeled C on the map in my previous post.

The drive up to Rosh HaNikra was great since it border the Mediterranean Sea.  In particular, we heard from a friend that it was a great site since the location has been not only a pivotal point communication and travel in recent history, but its significance can date back to Roman times.  Here’s a few pictures of the Mediterranean Sea, the Grottoes, and the border crossing between Lebanon and Israel at Rosh HaNikra.

In recent times, Rosh HaNikra had been used in World War II as a railroad station so that supplies could be easily transferred without having to go through the rugged mountainous terrain.  Unfortunately, as the war came to a close, tensions began to arise between Lebanon and Israel.  During the war of indepedence, Israel feared that Lebanon would use Rosh HaNikra as a junction to sneak troops into the country.  This lead to severing and blockcading the tunnel between the two countries and has been closed since then.  However, even before these times, Rosh HaNikra was used by the Romans to also transport supplies and the Grottoes shown above give tourists like us an amazing view of nature evolving at its finest.

Leaving Rosh HaNikra, we started to travel towards the historic city of Safed (Tsfat) now known for its Hasidic Jewish population as well as a central point for Kaballah..  For Shawn and I, we also had a personal interest in this city as it is not only where our advisor Allen Tannenbaum’s family is from, but its the name of our server here at Georgia Tech!  This is labeled as D on the map.  The city itself is not a very big tourist attraction unless you are concerned with Kaballah, but it does have a very historic feel to it. In fact, Shawn and I felt very much an outsider as most of the population of Safed are Hasidic Jews and we were arriving on Friday (the day before the Sabbath).  Nevertheless, the city had a very pleaseant and calm vibe, just don’t expect many people partying!

After eating lunch at Safed, we began our winding journey towards the tallest mountain in Israel, Mount Khermon.  This is labeled as E on the map.  Interestingly, this region of Israel, which is located right by Syria, is not recognized by the United Nations as apart of Israel. However, it is controlled by Israel and you will find Israeli troops and checkpoints near the mountain.  But before we reached Mt. Khermon, Shawn and I made the first driving mistsake of our trip as we missed a sharp turn to the mountain and ended up in a settlement called the Majdal al-Shams.  Luckily, a point of advice, when you get lost, always retrace your footsteps!  As we were leaving Majdal al-Shams, we met up with a local coffee shop owner who gave us directions to Mt. Khermon as well as free coffee and tea.  He also gave us a brief insight into the culture and everyday life, which is something that you definitely wouldn’t get on a tour.  Anyhow, here’s a few pictures of our drive up to Mt. Khermon, Shawn posing in front of the Ski office (it snowed the week before), and the mountain itself.  Note, on the otherside of the moutain is Syria and people are not allowed on the top of the mountainside since there are still some land mines.

At this point of the day, we realized that we only had about an hour or so before the sunset so we decided to quickly drive down to the Sea of Galilee and Tiberus to catch the view.  This is labeled F on the map and here are a few pictures of the historic biblical sea and city.

In fact, it was interesting to see that such a city that had such a significance impact on human history (as with the rest of Israel and the Middle East) had been turned into a tourist attraction.  While we did not see the city during the day, most of the people and natives said that the city of Tiberus was great for shopping and eating.  My particular take on the city was that the boardwalk on which we ate dinner was not nearly enticing as Haifa.  In fact, aside from the cool sunset view of Sea of Galilee, I would not recommend hanging out in the city for more than a few hours if that.  At any rate, we left the city when it was pitched black and headed back to Haifa to end Day 3.  The next post and final post will cover my visit to Jerusalem and our early morning hike up Masada to watch the sunrise.

The purpose of my trip was one motivated by a guest invitation to give a lecture about my on-going research in 3D visual tracking.  Of course, like traveling to conferences, it was also a time to meet fellow colleagues in the field and get a chance to visit an exotic place.  For now, I’ll just start summarizing the trip here and I’ll write a little more in-depth about the places we visited including pictures and opinions in the next post.

Day 1:

Jumping right in, my friend and colleague Shawn Lankton and I landed in Tel Aviv in the afternoon.  To our dismay, we didn’t realize how early the sun sets and how quickly it gets dark.  By the time we exited the airport, it felt like it was 9 PM (it was actually only 5 PM).  Nevertheless, we quickly rented our small and compact car, and were on our way to Haifa.  Being the “navigator” of the trip, I pulled out the maps and got familiar with the highway systems.  Here’s a quick google map look at the path we took all over Israel including the exact highways, toll roads, and places we visited during our stay.

View Larger Map

Becase we didn’t travel to all of these locations in one day, I’ll break up the trip by days beginning with our arrival in Tel Aviv. But before I do that, here’s a legend and short description of the places traveled (listed in order):

A:  Israel Airport. This is where we started. On the way back, we visisted Tel Aviv via highway 2.

B:  Haifa. This was where we were staying.  The Technion is located on the hills of Haifa, and it also allowed us to have a somewhat central position for visiting several places.  Each night except one night, we stayed in Haifa so a lot of the paths take us back here.

C:  Rosh HaNikra. This is the border between Lebanon and Israel where the Grottoes are located.

D:  Safed (Tsfat). This is a old city where Kabbalah originated from.

E:  Mt. Khermon. This is the tallest mountain in Israel and it happens to border Israel and Syria.  Interesting area since the U.N does not recognize this terrority as Israel, but is under Israeli control

F:  Tiberus and Sea of Galilee. These are historic areas that are attractive from both a tourist spot as well as historical biblical happenings.

G:  Jerusalem. Well, as our tour guide would put it, its the holiest of the … holies when it comes to cities.

H:  Ein Gedi and Masada. This was where we decided to float in the Dead Sea as well as hike up an old historic and relgious site of the Jewish people, which also happens to be located in the desert.  Absolutely beautiful.

I:  Haifa. This was where we were staying.

J:  Israel Airport. It was time to leave…

Not Listed.

Acre.  This is located on highway 4 going north to Rosh HaNikra.

Ceasera. This is located on highway 2 going south to Tel Aviv

Tel Aviv. We visited here on the way back to the airport.  Attraction as the biggest city in Israel.

Border Travels through Palestine. I was not the navigator here, but we followed our advisor through borders between Israel and Palestine.  We also passed through the area of where Amargeddon is suppose to happen (it was a peaceful agriculture field).  A bit hard to show on the map as these were smaller roads.

One thing that I will note here is that if you have ever traveled to Israel, it is quite fascinating how small the distances are from each of the major cities and even those cities that are located within Palestine.  Watching news, you sometimes get the feel that the areas are separated far apart, but in actuality, they are really close!  So close, that you can run from one city that is in Palestine to another located in Israel in less than 20 min (assuming there is no border patrol).  This was interesting for me since a major part of world conflict has to do with the state of Israel and Palestine.  Moreover, here we were driving on a highway and by using the naked eye, you could see the cities with relative ease as with neighbors in a suburb.

At any rate, since this was day 1, we finally reached the Technion and the local guesthouse.  Here’s a picture of the view from our balcony.

Day 2:

Waking up early from the mix of jet lag and with an early sun rise, Shawn and I explored the Technion.  Like that of Haifa, the university itself is built in levels on the mountainside.  The day consisted of working and meeting people and of course, we gave our lectures on the research projects that both Shawn and I are working on.  I have provided a preview here of my presentation as well as a downloadable link for those who are interested.  The main topic that I discuss is the framework of coupling 2D image segmentation and 3D pose estimation in a unified setting for the task of visual tracking.  Note:  Videos not included.

Download Link: Guest_Lecture_Presentation

Preview:

After the lectures, we ended walking by the beaches on the boardwalk of Haifa.  These beaches for me were one of the most relaxing and beautiful beaches that I went to, but perhaps this opinion is more slanted since I actually liked the city as a whole more so than others.  At any rate, let me conclude here.  In my next post, I’ll give a description of the places that we got to see and my novice travel tips for those that might be interested in visiting Israel soon!

Download Link:  Guest Lecture Invitation

Preview Invitation

At any rate, this time around, it dawned upon me of how engineering works even in the harshest conditions.  That is, whether we are playing music, or sending messages into deep space, coding theory and error correction is at the heart of designing the system.

For example, how can we transmit a message that is robust to transmision errors such as CD scratches? For now, we will assume our message consists of 1’s and 0’s.  The first step is called encoding.  In addition to just your message, encoding a message is done by appending what is known as parity bits in a unique fashion.  This results in a code word.  So why do this?

Well simply said, just because we would like to transmit a particular message over a channel, we might observe an entirely different message and code word at the receiver.  For those of you not familar with systems design, the transmitter can be considered a CD in our case and the reciever is your speakers.  The “channel” is the whole process of reading the data from the cd and interpreting it as music.  Now you can see that errors may be caused from scratches, dirt, and ware of your CD.  These parity bits that are discussed help and enable us to correct the underlying message and decode the correct data.

Lets dive right into an extremely popular coding scheme that is used, the 7-4 Hamming Code.  Here, our message is of length 4 bits (d₁d₂d₃d₄) with 3 appending parity bits (p₁p₂p₃), totally a code word of 7 bits (d₁d₂d₃d₄p₁p₂p₃).  Note, in reality, transmitted messages are much longer than our example since they are describing parts of data, i.e. track 5 of Pearl Jam “Black” on the Ten Album.  Let’s say that our message 1101.  How should we encode this?  Looking at the following figure should help us.

The 7-4 Hamming scheme encodes a message by making sure that if we input our data message into the above diagram at the specified locations, then the parity bits are added to ensure the sum of each circle is even.  Here I should note that we are using modulo 2 arithmetic for addition (e.g., 1+1 = 0, 1+0 = 1, …).  Doing so, we can then see the above message of 1101 is encoded as 1101100.  Hopefully this wasn’t bad!  Now, what happens if we transmitted this code word, but due to extraneous noise caused from our system, the message received is shown below.

We see here that our observed code word is 1111100.  The term “observed” here is important from a conceptual point of view.  That is, we have no idea of what the input code word is nor any of the system noise (we can only assume a general model for our system at best).  So how can we decode our message at the receiver with no knowledge of what the input is, but only that it was encoded via a 7-4 Hamming scheme?

One simple way is to put the code word in the above ven diagram, and see where the sum of each circle is not equal.  Doing so, one sees that the bottom two circles are not even in their sum.  But… should we flip the appropriate parity bits to be 1 rather than 0 or flip the third message bit from a 1 to a 0?  Either is a valid code word from a decoding point of view.  This means one can “correct” an error of the code word, but this code word may not be the underlying transmitted word.  This is known as a “decoding error.”  From this, the 7-4 Hamming code can be shown to correct at most 1 single bit error without resulting in a “decoding error.”  Unfortunately, there is no way around this and is dependent on the coding scheme that is used.  For the 7-4 Hamming code, if we have two or more bit errors in the transmission, the decoder may result in an error (but have a valid code). Because the decoder itself is beyond the scope of this post, I’ll refer you to this nice Wikipedia article. Also, given a coding scheme, different decoders may be used.  For example, if you use a convolution code (which can be found on my project’s page soon), one can use what is known as a Trellis Map to efficiently decode your message.  After we pass the information through the decoder, the correct code word will indeed be our transmitted code word.  While I have simplified this example and shown the use of the above ven diagram, more complex schemes will need a more powerful way of decoding involving parity check matrices (see wiki article).

Lastly, I will note that this coding scheme can only correct a single bit flip as opposed to several bits in a row, which is called a “burst error.”  So for our CD scratch example, where a scratch can be seen as several bits corrupted in a sequence, we will need to use other coding techniques in combination with the Hamming Code.

If you are interested, I have provided additional details on achieving a more efficient code via convolution codes.  This will soon be found on my projects page with MATLAB code for you to test out.  I have also written a short techncal summary that can be found on resume and publications page under “Technical Reports.  Here’s a quick link.

Every year, Glamour’s Top 10 college women competition honors leaders in every field (Martha Stewart won in 1961).  My cousin, who is a journalist, was one of the ten recipients for this year.

To say the least, she is an active member in her community, which is no suprise to me since she is very active in any type of family reunion!  In a field that is quite different than mine, she has been to many regions of the world to write about the atrocities that happen in everday life.  For example, she recently returned from Africa, where she was documenting HIV/AIDS. She also is a strong advocate for women’s rights in various cultures where women are deemed inferior.  Currently, she writes opinion columns for the Toronto Star, Canada’s largest newspaper.  Aside from this current award, her work has resulted in being the past recipient of the Top 20 Under 20 award, and in 2008 was named by the Globe and Mail and the Women’s Executive Network as one of Canada’s 100 Most Powerful Women. Of course, this isn’t her permanent job as she is now a junior at the University of Toronoto.  At any rate, here’s a photo I took of her in the magazine, and if its hard to read, here’s a link featuring her as well as the other top 10 college women honored by Glamour Magaizine.

If you are interested in her work, I suggest you check out her blog here.  Keep up the good work Jasmeet!

Journal Checklist:

1) Create a separate submission folder:
This seems somewhat trivially, but my main point here is to separate the manuscript.pdf from being editable. I have found that when I begin the submission process, I always am making minor grammatical or sentence style changes (e.g., Use “Moreover” vs. “In Addition”). There’s always something I can correct. My advice, if you get to this point, then your manuscript is ready to submit. In order from you to keep changing these details, make a submission folder and place that paper there and as the saying goes, “don’t look back.” That is until you finally get to the “approval of all documents” stage where you can view your submission online.

2) Create a cover letter:
I believe for most journals, you will need to create a cover letter identifying the paper’s title (and perhaps author list) as well as the main idea and contribution. This is pretty standard and can be likened to a cover sheet. Here’s a quick example of a cover letter that I just submitted.

Cover Letter

Please note that this is general, and check your journal guidelines to see if more information is needed.

3) Create a shortened abstract:
Now you might be able to skip this step depending on your paper, but each time that I have submitted a paper, whether it was to a journal or conference, I realized that my abstract went over the “word limit.” Ideally, an abstract should be the shortened version of the paper, but then again, it might be over by 30 words. So without distracting from the main idea, make sure that your abstract is short enough to submit, otherwise doing this on the fly can be sometimes distracting, let alone annoying!

4) Gather keywords describing your paper:
Keywords are important. It might seem that this is a minor detail in the submission process, but those few words that you choose matter alot! For example, the chosen keywords might play a factor on how the associate editor will distribute the paper amongst reviewers. Moreover, if your paper is to be accepted, other researchers in your area will usually perform a general search and these keywords aid in placing your paper corresponding to one’s search. So, take some time and choose a few sub areas that is representive of your work. This pretty much goes for conference papers too.

5) Create a summary of changes (if similar conference paper exists):
Usually, my collaborators and I tend to write to conferences before we submit to journals. The reason being is that we would like to get our preliminary results of the work out to the community, and it sets the foundation for the more important paper, its journal version (of course going on vacation to a nice place is not a bad consquence). Then again, if you decide to stick with research, I have seen some of my colleagues that prefer to just submit to journals. Nevertheless, should your journal submission be a more in depth and detail work of a similar conference paper, most journals ask for a “summary of changes.” This short one page or less document is simply a description of the extensions and changes or differences from your conference work. For example, your extensions might be to broaden a certain framework in a theoretical setting and to include more experiments. Basically, the journal paper is seen to be the more “complete” work, but you need to have differences between the two works and they must be documented and presented. Here’s my recent “summary of changes” example:

Journal and Conference Summary of Differences

One thing that I will note about this document is to be honest! I have found that some people tend to exaggerate the differences between the two papers. My advice is don’t. Be forthcoming about your research, even showing both the success and failures. I was once told that at most a person would be lucky to have a total of 3 to 5 great ideas in their lifetime. So if you feel strong about your work, let it speak for itself. Being on the side of a reviewer, I know that one thing that will annoy him or her when critique’n your paper is if you exagerate the actual accomplishments. It might sound naive, but your reviewers are human beings too and they will appreciate the full assesement of your contribution.

6) Prepare a list of reviewers:
Most IEEE journals allow you to have a “preferred” and “non-preferred” list of reviewers and associate editors. This is up to you and is not necessary to fill out, but there might be a research group or a few people who might better understand your work. You can suggest these people during the submission process so keep this in mind.

7) Submission and Copyright form:
After you proof read your current submission that is the database, its the dreaded “submit” button that everyone always thinks at least twice before hitting. However, if you are submitting to an IEEE journal, you will find out that the process does not stop here just yet, which leads to more confusion! You have one extra step, which is the online copyright form. Don’t worry, its extremely simple. All you have to do is follow the instructions and type in your name that you agree to the rights and guidelines. This beats printing the form, signing it, and faxing it over.

Hopefully, this will help you when it is time to submit your journal. In the future, I will try to discuss major revisions, minor revisions, and final acceptance submission associated with the whole journal process.

I’ve found one of the more tangible examples is path planning by GPS. Of course, you may all know how many miles it roughly takes you to go from your home to your favourite restaurant. The more interesting question is how exactly does GPS compute the mileage?  The obvious answer here is the miles that it takes you while traveling on a road or other legal pathways (e.g., not through the local shopping mall).  It takes me 7.6 miles to go from Georgia Tech to my temporary favorite, Panera Bread.  For fun, here’s a quick view of the path.

View Larger Map

Ok so with that trivial example, the next question that begs to ask is well what if it isn’t distance between two points in a city, but between two cities in different continents? Better yet, what if you decided to fly a plane rather than drive — then you can fly right over that shopping mall!  However, if you have ever flown in an airplane and have seen a flight of path that is being displayed, you realize that the path is not a straight line, but rather it resembles an arc (even assuming times where turbulence is being avoided).  Realizing that airlines what to be the most efficient, it would seem that a straight line would be better?? The short answer is no. In fact, the arc pathway that is taken is the shortest distance possible between the two cities. Below is flight path (in red) from Atlanta Hartsfield Airport to this year’s Computer Vision and Pattern Regconition Conference (CVPR) sight in San Francisco.  Note, the black line is a straight line for reference.

The question is why? Well, it has to do with the surface that we live — the earth or better yet, a sphere! Tthe shortest distance between any two points that live on a sphere can be characterized by great circles or arcs.  In differential geometry, this shortest distance is known as a geodesic.  The first inclination is then why do our GPS compute distance via miles traveled on roads, doesn’t it make more sense to compute the arc between your home and that favorite restaurant of yours?  First lets assume that the resturant is in San Franciso and you are in Atlanta.  Then the answer is yes and no.   If you had a plane, and were flying by plane, then yes.  This is because you can assume that the surface that you travel on has no restrictions except that it is a sphere.  Now if you are traveling by road, then you can think of your surface as not really being a sphere.  The buildings and restricted legal pathways make the surface on which you compute distance an entirely different one!  Thus, the GPS computes this distance, in which is interestingly referred to as the “taxicab distance” or “taxicab geometry.” The arguement of “shortest path” for path planning by GPS is beyond our present discussion. If you are interested though, the area has been studied in many control problems. Here’s a link to the webpage of Evdokia Nikolova, who gave an excellent seminar here at Georgia Tech.

Now lets take what we have learned a step further.  Say that I have a set of data points that characterize  specific object p and another set of points that characterize object q, what is the distance between these p and q?  The first question that you may ask is what surface do these points lie on.  If they live in the Euclidean world, then the answer is simple.  Its the distance that we have learned since we were little, and one that Shaq has once used to describe his skills as an athlete:

Shaq:  I’m like the pythagorean theorem. Not too many people know the answer to my game.

Unfortunately, to the shagrin of Shaq, this distance,which is also known as the Euclidean distance, only answer’s the problem for linearly related points or those points that live on a Euclidean surface.  If these points were derived from a probability space (e.g., p and q now represent probability density functions) the answer is not so simple.  In fact, this now leads us to a new class of distance metrics that try to measure the similarity between two probability density functions.

In computer vision and segmentation, being able to characterize an image through statistical measures has recently grabbed many people’s attention.  Segmentation, which is a low level vision task, is the idea of being able partition a certain object from that background.  You can imagine now that in the above examples, p now represents an object we would like to grab in an image while q is the background.  Interestingly, the ability to quanitify the distance between the two probability measures plays a major role on how effective one’s segmentation algorithm is.  This has become a major research of mine, where it has been my joy to work with Tryphon Georgiou to introduce a new way of characterizing distance in the area of imaging and perhaps, even in your everday life! Please check out my project’s page for more details.

Contact Info:

Romeil Sandhu
Georgia Institute of Technology
Department of Electrical and Computer Engineering
MINERVA Research Group
313 Ferst Drive, Atlanta, GA, 30318
rsandhu@gatech.edu