Speech:Spring 2013 Marc Southard Log


 * Home
 * Semesters
 * Spring 2013
 * Proposal
 * Report

Week Ending February 5th, 2013
https://basecamp.com/2144088
 * Task:

1. Date created a date picker (jqueryUI datepicker?) with current date as current day.

2. Ability to edit experiements (create an edit.php? With similar look to add.php but just edits).

3. Fix description not showing in view page.

February 5, 2013
 * Results:

Spent some time coming up with a new way to use the datepicker. Simplified it so that all you need to do for any input on the site is add class="datepick" and the plugin will function correctly. This is a move off of the default id="datepicker" which has some problems scaling.

February 4, 2013

After working with Scott and the current login page, datepicker should be working in production (add.php)

Got the theme for the datepicker working on the test page.

http://i.imgur.com/0iv4HL5.png (flick)

February 3, 2013

Got the Datepicker working for jQuery on my test page. I will upload the changes tomorrow. This will also allow for easier future use of more jQuery goodness.

http://jqueryui.com/datepicker/

February 2, 2013

Something is not working quite right with XAMMP. TortoiseSVN is up and running but PHP pages are not. Will redownload/reinstall and see where I'm at.

Get a jquery calender plugin implemented. Adding a new page that will allow you to edit existing experiments. Then the description field when viewing experiments currently doesn't show correctly, it looks like it is setup correctly so finding out why will be the most time consuming. Getting jquery working, hopefully it won't be much of an issue. I also haven't dealt with SQL+PHP so getting description to display as it should might be a bit difficult at first, this goes for setting up an edit page as well.
 * Plan:
 * Concerns:

Week Ending February 12, 2013

 * Task:

-Fix Description

-Edit Experiments up and running

-Datepicker added to search page

February 12, 2013
 * Results:

Getting the edit page fully functional. Most data is showing up, just need to work on the rest of it.

February 11, 2013

Read logs, working on setting up MySQL on my home desktop to bounce the SpEAK queries off of.

February 9, 2013

Read logs

February 6, 2013

Created a skeleton page for editing. Passed the result to Scott to hash out the PHP-SQL for our dbController.


 * Plan:

Work on the edit page so by next week it will be functional.
 * Concerns:

I think the biggest problem was setting up the correct MySQL queries which I made a point of hashing out in class. Might revisit on Skype if need be.

Week Ending February 19, 2013
-Catch up with Scott and Nick.
 * Task:

-Sync MySQL with the live tables on stem.

-Get edit up and functioning being the most important part.

February 19, 2013
 * Results:

Talked with Tyler about what his group has been working on.

Worked with Scott to get the correct MySQL setup of SpEAK running.

Updated the dbExperiment controller with an UPDATE function for view.php

Added update.php as a file to handle when an experiment gets modified.

February 18, 2013

Fixing the description field when viewing an experiment. Scheduled a time to meet with Scott for help with MySQL Tuesday. Setup LogMeIn on my home desktop for this purpose.

February 17, 2013

Read logs

February 14, 2013

Spent time catching up with any updates other team posted on the logs. Also talked with Scott about what went on in class Wednesday.

Finish the edit page. Keep up with reading others logs. Wasn't able to meet with my team in class, just want to make sure I'm on the same page
 * Plan:
 * Concerns:

Week Ending February 26, 2013
-Work on the 'all' experiments page.
 * Task:

-Implement a data table layout.

February 26, 2013
 * Results:

Working on getting data table plugin running on the SpEAK site. Proving extremely difficult. Read logs.

Finally got the data table working. All files have been added to the repository including:
 * /javascript/dataTable.js
 * /javascript/jquery.dataTables.min.js
 * /javascript/complete.min.js
 * /css/demo_table_jui.css
 * /includes/head.php [modified]
 * experiment.php

February 25, 2013

Got the data table plugin up and running locally. Next step is integrating it with MySQL on the new experiments page.

February 24, 2013

Read Logs.

February 23, 2013

Started working on the experiments page.


 * Plan:

Setup a new page in SpEAK that can display all experiments.
 * Concerns:

Haven't setup the data table I plan on using before.

Week Ending March 5, 2013
Update the header to include the new experiment.php page.
 * Task:

Fix/Comment code.

March 2, 2013
 * Results:


 * Commented code within SpEAK to clue in future coders.
 * Cleaned/Refined current code on experiment to make it easier to read.

March 2, 2013

Read Logs

February 28, 2013

Read Logs

February 27, 2013


 * Talked with Scott about SpEAK and the progress we've made during the last week. Also learned more about the progress Nick has made with uploading files.
 * Updated head.php to link to experiment.
 * Might be worth rewording some of it as well now that we're adding new content.


 * Plan:

Updating the site for the new page I created. I want to comment more code to make it easier to follow.
 * Concerns:

None at the moment.

Week Ending March 12, 2013

 * Task:
 * Make the experiment page the new index page.
 * Display experiments by date created desc.
 * Update the header menu with appropriate changes.


 * Results:

March 12, 2013
 * Moved experiment.php to index.php as the new home page.
 * Deleted the moved experiment.php page as it's no longer needed.
 * Updated the /include/head.php to reflect the removal of experiment.php
 * Updated /controllers/dbExperiment.php so that it is no longer limited to grabbing the first 25 results. Rather, the index page will paginate every 25 results going forward.

March 10, 2013
 * Read logs.

March 8, 2013
 * Read logs.

March 6, 2013
 * Helped Scott debug the site.
 * Worked on rewording the menu with better descriptions.


 * Plan:

Execute what was discussed in class. Outlined in task.


 * Concerns:

Week Ending March 26, 2013

 * Task:


 * Meet with Group C on Google+ hangout Friday @1PM
 * Contact Mike Mailloux and fill him in on the details.
 * Get an understanding of setting up Experiments.


 * Results:

March 21, 2013


 * Contacted Mike via email, told him about the Google+ hangout that Josh setup.
 * Started reading up on the experiment setup notes & speech project notes.

March 22, 2013


 * Met with Josh & Harry on a Google+ hangout. Josh went through the setup routine for getting an experiment running.
 * I'm going to set aside some time to try and get an experiment running before the next class.

March 24, 2013


 * Read logs

March 26, 2013


 * Read logs


 * Plan:


 * Run through the basics of setting up an experiment. Attempting to successfully complete the task.
 * Concerns:


 * Finding time to meet with the other group members. The more people in the group, the more difficult that can be.

Week Ending April 2, 2013
Get an experiment ready for decoding.
 * Task:


 * Results:

March 27, 2013
 * refer to http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_3rd_Group_C

March 28, 2013
 * Read logs

March 31, 2013
 * Read logs

April 1, 2013
 * refer to http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_3rd_Group_C

April 2, 2013
 * refer to http://foss.unh.edu/projects/index.php?title=Speech:Exps_0042&action=edit&redlink=1
 * refer to http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_3rd_Group_C#Group_Log


 * Plan:

By next week I want to run through the entire setup process for an experiment.
 * Concerns:

Miraculix is problematic at the moment. So I will be working on Obelix but I would like to run a second experiment on Miraculix if possible soon.

Week Ending April 9, 2013

 * Task:


 * Create a Language Model, decode & finally score the train to complete the experiment.
 * Setup a section of the dictionary that I'll be assigned during the week.


 * Results:

April 3, 2013


 * Created a Language Model.
 * Decoded the experiment.
 * Ran the score successfully.
 * refer to http://foss.unh.edu/projects/index.php?title=Speech:Exps_0042&action=edit&redlink=1
 * refer to http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_3rd_Group_C#Group_Log

April 4, 2013
 * Read the Capstone_GroupB&C notes. Staying up to date with how the project leaders are coming along with the 5 hours of data. Will have to see how the dictionary work is divided up.

April 7, 2013
 * Capstone Group B&C notes were updated.
 * Looked up what words needed phonetics and picked a section that I thought others had not chose yet.
 * Sent Eric an email with the phonetics for letters (E) (F) & (G).
 * Read the rest of the notes, the plan/outline for execution.

April 8, 2013
 * Read logs

April 9, 2013
 * Read more logs


 * Plan:


 * Finish running a full experiment.
 * Setup my portion of the dictionary for the section of the alphabet I get assigned.


 * Concerns:


 * Time to work. Also working while others are may cause data to be lost. That would be frustrating.

Week Ending April 16, 2013

 * Task:


 * Read the errors logs for the first 5 hour train.
 * Try to reduce the error rate for the latest experiments.


 * Results:

April 11, 2013
 * Checked the Capstone GroupB&C documents for this week.

April 12, 2013
 * For this week our accuracy for the tail-end of the 5 hour corpus was 33.1% (down from in the in class amount).
 * We're running a repeat of experiment 0075 to see the results for when the dictionary doesn't use stress indicators.

April 15, 2013
 * Group log: http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_10th_Group_BC#Group_Log
 * Ran into some problems with the Gen Feat step I took for a responsibility. Documented the error and reported it to Eric. I think someone may have completed this step and just not left a note.
 * We should be ready to run the train for experiment 0083 now finishing our task for this week...


 * Got sphinx_train.cfg fixed in the etc folder. There were some errors that might have been left from an mistaken write to the file. Extra characters throughout the file, weird!
 * Generate feats took about 3 minutes to run, and completed without errors.

April 16, 2013
 * Read group documents, group logs, individual logs.
 * We finished 4 5-hour experiments.
 * Changes to the corpus are still somewhat inconclusive.
 * Word accuracy is still fluctuating without any real indicator as to what works the best for improving.


 * Plan:


 * Communicate with the group to stay up to date with our progress.
 * Concerns:

Week Ending April 23, 2013

 * Task:
 * Poster completion
 * Experiment accuracy improvement


 * Results:

04/20/2013
 * Completed the poster Wednesday

04/21/2013
 * Read Logs

04/23/2013
 * Current experiment went a little bit better. Our group almost hit a 31% this time.
 * Read logs


 * Plan:


 * Complete the poster for the SpEAK work done this semester.
 * Try to hit a better percentage for experiments.
 * Concerns:

Week Ending April 30, 2013

 * Task:
 * Compare the results of the older experiments against the changes made over the past week to verify if improvements were made.
 * Work will be divided up between training and debugging.
 * Debugging work will rely mostly on individuals documenting errors to see what the main problems with accuracy seem to be.


 * Results:

04/24/2013
 * Set up the task/plan for the week.
 * Our team will be working on improving experiments as we reached an accuracy rating of 31.9% over the past week.

04/25/2013
 * Read logs.
 * Group http://foss.unh.edu/projects/index.php/Speech:Spring_2013_April_10th_Group_BC

04/29/2013
 * So, I am going through the logs catching up, I just spent this weekend moving and had no computer to work from, since Friday night until...now. Anyways, Eric setup tasks for the group this week and I will spend time as I can working on what is left. Was simply a bad week in terms of having any free time between classes.


 * Plan:


 * Eric/Josh as the team leaders will find out what work they want to give out to everyone else. Upon doing so the group will complete tasks as we can.
 * Concerns:

Week Ending May 7, 2013

 * Task:
 * Group B&C will be running experiments 98/99/100


 * Results:

05/04/2013
 * Read the plan for this next week.
 * Brian did work on experiment 98 to get it rolling, started where he left off.
 * Generated the phone list for experiment 98.
 * Generated the Feats data as well for experiment 98.
 * 0098: Test on Train (0089) using last_5hr/train
 * 0099: last_5hr train using genTrans6.pl
 * 0100: Test on train 0099

05/05/2013
 * Read Logs

05/06/2013
 * Read Logs

05/07/2013
 * The final experiments 98/99/100 are complete.
 * Evals tomorrow.


 * Plan:


 * Concerns: