Speech:Spring 2013 Nicholas Regan Log


 * Home
 * Semesters
 * Spring 2013
 * Proposal
 * Report

Week Ending February 5th, 2013

 * Task:
 * Author who is logged in will be default as the editor
 * Add image attachments to the page

Thursday Jan 31- Looked at previous semesters and what they have done and went over SpEAK and what we got done and what we did't get done last semester
 * Results:

Monday Feb - Set up Xammp on my computer and looked around the web to find out how to add a file to a web page.


 * Plan:

My plan is to get Tortes SVN up and running with Google Code, after that get it running on my computer at home. Once it is up and running attempting to get the Author to auto fill-in, as well as give the page the ability to add a file to it.


 * Concerns:

My main concern is getting the upload to work because I have never done something of this sort before.

Week Ending February 12, 2013
Add user by default to the Add page Make the site have the ability to upload and store files.
 * Task:

Wednesday- Got xammp running and working locally on my computer. Monday- Looked over others Blog posts from the past two weeks.
 * Results:

Tuesday- looked into the ability to upload a file to the site

The plan is to implement the ability to upload a file tomorrow durring or after our meeting to have it complete for our three week goal.
 * Plan:
 * Concerns:

Week Ending February 19, 2013

 * Task:

Wednesday 2/13 - I got the ADD page to auto populate the User who is logged in to add to the ADD page. I helped look at and try and figure out the issue with Caesar until 4.
 * Results:

Saturday 2/16 - I read others post on the Wiki via Google Reader

Sunday - 2/17 I read thought others logs on what they have been doing this week. To get the recent changes to show up on Google reader 1. Simply goto http://www.google.com/reader 2. click subscribe 3. In the box that comes up put http://foss.unh.edu/projects/index.php?title=Special:RecentChanges&feed=atom in the box 4. Click Add and every time there is a change to a persons page you can view it in one page.

Tuesday - 2/19 Looked into adding files to the the database, the concern I have with doing this is being able to change the upload location from the temporary store location that mySQL will save it in. I think their is an easy solution to this I just need to find it.

youtube video This video has a good tutorial on how to do the changing of the file storage location. dhost This link lays out some of the same guidelines for uploading a file.

Tomorrow before class I plan on implementing the add file link to the add experiment page.
 * Plan:
 * Concerns:

Does anyone know the max sizes of the files we are going to be uploading in the future to the SpEAK site?

Week Ending February 26, 2013

 * Task:

This week my tasks are to backtrack and update our group wiki with what has been done over the semester. I am going to look at our Google code repository and with that I am going to just enter the changes that have been documented. This week I am also going to add the ability to upload any files to the server so users can attach a file to the experiment page.

2/21 I updated our groups media wiki page to have the changes that have been made to the speak site shown on our wiki.
 * Results:

2/23 Read longs

2/24 Working on adding an option to our site to have the ability to upload files.

2/25 Read logs

2/26 Read a few different website on how to get the upload function to work


 * Plan:


 * Concerns:

My main and only concern right now is getting this to work, I have spent some significant amount of time trying to figure it out and I just can't wrap it all together and implement it.

Week Ending March 5, 2013
Finish implementing the upload function
 * Task:

get some feedback on users of the page
 * Results:

2/21 Read Logs

3/5 Read Logs


 * Plan:


 * Concerns:

Can some people goto SpEAK and leave some feedback HERE. Thanks

UPDATE!!! Login ID: admin password: nimda

Week Ending March 12, 2013

 * Task:


 * Results:

Friday March 8th read logs

Monday March 18th read logs

wednesday march 20 updated the experiment group wiki


 * Plan:


 * Concerns:

Week Ending March 26, 2013
Got assigned to B Group with Eric as our leader, myself and Mike Brown and Brian Drouin

My tasks for this week are to
 * Task:

1. Create a unique experiment number for myself

2. run a decoder

3. run a scorer

Saturday March 23rd, Read logs
 * Results:

Sunday March 24th, Read emails from Eric and Read logs

Monday March 25th, Logged into Miraculix because Obelix doesn't have Sox installed as well as SC lite isn't working properly on it. I set up the acoustic training for my experiment 35 edited the config file and it is taking a while so I am calling it a night and hope it is done recording so when I come back on it is done recording, so I can continue moving towards a train. This process finished, I continued on with the directions and created the scripts and its associated audio-file list. Mike Tromos and I had a Google hangout for a few hours working on this together, for a couple of hours and when we would get stuck we would help each other other, it was very helpful to have to brains working to fix a problem either of us would run into.

Tuesday March 26th, Got an email from Eric saying he say my error on the Wiki and told me to remove all the files I have in my 0035 folder and start from scratch because "There is no transcript (0035_train.trans) in your 0035/etc folder. Thus causing the dictionary process to fail, thus ultimately causing the genPhones.csh script to fail." So I deleted the files and am going to start from scratch. I successfully loaded the dictionary again, and now I am getting an error at the same spot but a different error.

ERROR

miraculix Exp/0035> /mnt/main/scripts/train/scripts_pl/make_feats.pl -ctl /mnt/main/Exp/0     035/etc/0035_train.fileids -cfg not specified, using the default ./etc/sphinx_train.cfg -param not specified, using the default ./etc/feat.params mkdir /feat: Permission denied at /mnt/main/scripts/train/scripts_pl/make_feats.pl line 103

Line 103 mkpath(catdir($ST::CFG_FEATFILES_DIR, $dir));

My plan is to follow the instructions that Eric has sent out to the group for how to set up our own experiment number and then run a train on our server. The server I am going to be using is Miraculix I will be using Putty to SSSH in. I haven't been working with this all semester so its my first dive into this aspect of the product so I have no idea what to expect.
 * Plan:
 * Concerns:

Error when trying to generate phone list

miraculix 0035/etc> ./genPhones.csh 0035 0035 awk: cmd. line:1: fatal: cannot open file `0035.dic' for reading (No such file or directory)

miraculix 0035/etc> ls 0035 0035.filler  0035.phone  feat.params  genPhones.csh  sphinx_train.cfg

Anyone have any idea whats going on here???

Week Ending April 2, 2013

 * Task:

Run a mimi.eval train on experiment 0045


 * Results:

Wednesday Mar 27th- created experiment 0045 to see if that was going to make the error I was having go away. My server doesnt have sox so I am using verleihnix.

Thursday Mar 28th Was able to get over the problem I was having yesterday. I was able to Generate the phone list, as well as generate Feats data. I was able to successfully run a train, I did run into the issue 1 and was able to add the missing words to the dictionary. Eric showed me how to copy a file in someone else s dictionary into my own experiment that way I didn't need to create my own from scratch. Created the model for experiment 0045, that was pretty straight forward. At this point I have made the DECODE directory,and am running the decode on the experiment.

Saturday March 30th Read the crazy amount of log entries

Sunday march 31st logged in to see if my decode had ran and it had. I went to use sc lite to score that file and got as expected

Error: Not enough Reference files loaded Missing: (sw2466b-ms98-a-0071 )

A BRIDGE TO FOREVER THOSE ARE REALLY REALLY MIND EXPANDING LAUGHTER BOOKS HIS CONCEPTS ARE SO DIFFERENT THAN WHAT I WOULD HAVE EVER DREAMED OF (sw2466B-ms98-a-0071)

(sw2470a-ms98-a-0014)

A BUTTONS PC FILE PROGRAM THAT UH IS A VERY SIMPLISTIC RELATIONAL DATA BASE FOR LABELS AND THINGS SO I GUESS I WOULD HAVE TO SAY ON A HOME SETTING I'M PROBABLY ON (sw2470A-ms98-a-0014)

(sw2479a-ms98-a-0071)

RIGHT SHE GOES TO SEMINARS AND UH SHE GETS HOME VISITATIONS BY THE STATE UH UH I DON'T KNOW THE STATE BOARDS I GUESS SOME OF THEM AND THEN SOME BY THE ASSOCIATION AND THEY (sw2479A-ms98-a-0071) (sw2493b-ms98-a-0059)

WELL THAT'S HOW THE SHOW THAT'S HOW THE SHOW ENDED IT JUST LEFT YOU HANGING THERE LIKE YOU KNOW LIKE YOU SET ME UP YOU FED ME ALL THIS INFORMATION INFORMATION YOU KNOW THAT MADE IT LOOK LIKE THIS OTHER PERSON WAS GUILTY YOU KNOW AND (sw2493B-ms98-a-0059)

(sw2534a-ms98-a-0098)

THEY HAVE THE GLOVES ON THE PAWS AND THEY'RE THEY'RE USUALLY TWO COLORED UH BUT THIS ONE IS UH (sw2534A-ms98-a-0098)

(sw2547a-ms98-a-0059)

(sw2547b-ms98-a-0099)

NO WE ARE WEST SEVREN IS EAST NEAR THE EASTERN SHORE IF I HAD (sw2547B-ms98-a-0099)a-0059

Monday 4/1 I was able to complete the scoring process and got results, I updated my experiment 0035 and 0045 with the template that Eric had sent to us to keep the experiments wiki unified. I helped Mike T get his train running. My plan for this week is to meet with Eric at least once on Google Hangouts and then hopefully complete the whole thing then. If not I know that the wiki has very detailed steps on what we need to do, as well as error fixes for almost all steps that we need to fix the common problems. My main concern for this week is defiantly not being able to get this train and decode to work properly in time.
 * Plan:
 * Concerns:

Week Ending April 9, 2013

 * Task:

Saturday April 6th Read Logs Via Google Reader

Sunday April 7th Read Logs Via Google Reader

Monday April 8th Added words to the phonetic dictionary so we can get started on the train soon,

Tuesday April 9th


 * Results:


 * Plan:


 * Concerns:

Week Ending April 16, 2013

 * Task:


 * Results:

Saturday April 13th- read logs

Sunday April 14th- Read Eric's emial's about our experiment 80 not being able to decode our script. So we are now going to switch over to experiment 82. As well I read through the Logs to see what my other teammates have been up to.

Monday April 15th- I logged into verleihnix and was going to do the scoring of the data and the decode hasn't been completed. So once that has been competed I will hop on and complete the scoring for experiment 0080.

Tuesday April 16th- I worked on editing the experiments wiki pages, groups BC completed experiment 0080-0083. I scored experiment 0083 and we got an alignment issue when scored and a terribly high error rate. We are going to have to work on that for the next assignment, figuring out why our alignment on the most current experiment we ran. We were trying to improve our word accuracy rate and it has for some reason just gotten noticeably worse.


 * Plan:

My plan is to do one of the steps that Eric has set up for us to complete before weeks end tomorrow Monday the 15th.


 * Concerns:

Week Ending April 23, 2013

 * Task:


 * Results:


 * Plan:


 * Concerns:

Week Ending April 30, 2013

 * Task:

Friday April 26th- Spent three hours at the URC presenting the poster for my SpEAK group.
 * Results:

Sunday April 28th- I read the email Eric sent out and read logs from the past few days.


 * Plan:


 * Concerns:

Week Ending May 7, 2013
Update logs for experiments we have done
 * Task:


 * Results:

Wednesday May 1st- Updated experiment logs for the three experiments we did the previous week Thursday May 2nd- Read logs on Google Reader Sunday May 5th- Read logs on Google Reader


 * Plan:


 * Concerns: