Speech:Spring 2013 Vincent Gagliardi Log


 * Home
 * Semesters
 * Spring 2013
 * Proposal
 * Report

Week Ending February 5th, 2013
Orchestrate with all the different groups to find out their goals for both team work and individual work.
 * Task:

1/31 - Compiled a list of everyone's email for contact purposes copied all of the team info.
 * Results:

2/3 - Read logs for other team members

2/4 - Re-read proposals for the past semester's groups to gain a better idea of what is to be done for our proposal.

Communicate with each team in class on 2/6 to get a general idea for everyone's plan. Identify individual/team responsibilities and descriptions to apply to the proposal.
 * Plan:

My ultimate concern is having everyone on the same page when it comes to the proposal and making sure everyone is satisfied.
 * Concerns:

Week Ending February 12, 2013
Create the draft of the proposal.
 * Task:

2/6 - Create the bare-bones for the proposal on the wiki. The wiki is broken up into each team and specific tasks the plan to accomplish. Talked to each group and asked for a general description for what each plans hopes to attempts. Got a general idea for the first 4-week timeline.
 * Results:

2/11 - Read logs

2/12 - Read logs

Comment from Mike B. : My portion of the systems group proposal is up on the proposal wiki.


 * Plan:


 * Concerns:

Week Ending February 19, 2013

 * Task:

2/16 - Emailed the groups asking for feedback regarding the first 4 week timeline.

2/17 - Read Logs

2/18 - Read Logs

2/19 - Didn't receive much feedback regarding the other teams' goals. Read through more logs to try and determine each group's track.

As of currently we have a rough draft of the proposal. There are still some holes to fill regarding the specific areas of each group. Some timelines are being put together but most are pending feedback from the groups.
 * Results:

Edit the draft proposal to be more specific. Talk with each group about their desired goals for the coming weeks. Generate 4 week timelines based on each group's goals and also have discussions about future 4-week goals.
 * Plan:

Not getting much feedback via E-mail.
 * Concerns:

Week Ending February 26, 2013

 * Task:

2/23 - Read logs

2/24 - Read logs, reviewed changes to wiki

2/25 - Reviewed more changes to the wiki. Outlined all goals to keep track.

2/26 - Provided minor updates to the wiki. Plans to have the timeline/goals in a uniform format.

We got good feedback from the groups regarding each teams goals throughout the weeks. Added the wiki into the timeline and the proposal is well filled out now.
 * Results:

Set plans to create new goals for the next 4 week timeline and add uniform formatting to all the timelines.
 * Plan:

Getting the second set of goals and seeing how the first set of goals is met. Need to provide more updates to the general information in the proposal.
 * Concerns:

Week Ending March 5, 2013
2/4 - Read logs
 * Task:

2/5 - Read Logs

Kept up with what other groups were doing based on wiki changes and logs.
 * Results:

Work with the Data Capture group for the poster.
 * Plan:


 * Concerns:

Week Ending March 12, 2013

 * Task:


 * Results:


 * Plan:


 * Concerns:

Week Ending March 26, 2013
In the previous class we got new teams (TEAM A) to continue to run trains and model the data. We discussed different approaches for this in class and ultimately decided to meet online in a Google Hangout to collectively run a train to get everyone on the same page.
 * Task:

2/24 - Group A had a group meeting on Google Hangout to go over the process of running a train. Used the guide on the Wiki on how to run the train. We ran into a few errors, particularly for myself trying to connect to a machine as I was not in class the day we set them up.
 * Results:

2/25 - Read Logs

2/26 - Scott was able to fix my problem regarding connecting to a machine and now I can do so. I am currently using the Miraculix machine to attempt to run a train since some of the other machines seem to have miscellaneous problems when working with the trains.

As a group we plan to meet in class on Wednesday to model and score the information we obtained as a group so everyone is familiar with the process.
 * Plan:

Not being able to connect to a machine. - Resolved
 * Concerns:

Week Ending April 2, 2013
The task for our group this week is to collectively be able to create the language model and decode. We will be meeting online again on Sunday to go over this as a group. Hopefully by this point everyone in the group will at least have been able to run a train.
 * Task:


 * Results:

3/30 - Ran a train in Exp 0040 using the mini corpus (mini/train). Noticed that one step of the train had over 11,000 errors.

3/31 - Met with the group online in a google hangout. Created the language model and decoded. Some errors came up when attempting to scores (Solutions found in wiki).

4/1 - Read logs

4/2 - Attempted to fix the errors when trying to score and was unable to do so. Plan to meet in class tomorrow to discuss the process.

Have the group be able to run a train. From there, be able to create the language model and decode. Meet online on 3/31. Questioning whether the train ran correctly - 11,000 errors in one step.
 * Plan:
 * Concerns:

Week Ending April 9, 2013

 * Task:

Collaborate as a group to make sure everyone can run a train, decode, etc. Also work with team D on getting the transcripts for the first 5 hours and creating the experiment.


 * Results:

4/7 - Read logs

4/8 - Read logs

4/9 - Worked on creating another experiment (0079) and attempted to create another experiment with the first 5 hours. I seem to only have access to Miraculix so I will have to change my authorized_keys tomorrow in class.


 * Plan:

Meet with the group to discuss further plans for more experiments.


 * Concerns:

Only having access to Miraculix.

Week Ending April 16, 2013

 * Task:

Collaborate with group D and work with gentrans to generate more accurate transcripts.


 * Results:

4/10 - Meet with group in class. Went over everything that needed to be done, including new experiments and work with gentrans.

4/11 - Read logs

4/15 - Google hangout with group A/D members. Create new experiment 0084.


 * Plan:

Create more experiments with lower word error rate.
 * Concerns:

Communication within the group, having many members in the group.

Week Ending April 23, 2013

 * Task:

Create experiments on the first five hour train.


 * Results:

4/17 - Met in class with the group. Agreed upon doing more experiments with the first five hour train and updated gentrans script. Hopefully get results with lower word error rate.

4/18 - Read logs. Communicated updates via email thread with the whole group.


 * Plan:

Decide upon contiuned experiments for the final report based on the results of AD vs BC.


 * Concerns:

Week Ending April 30, 2013

 * Task:

Create the final report for the total progress for the whole project. Divide each section up to be written by members of AD with support from Erik based on BC's experiments.


 * Results:

4/24 - Created a shared document with the group in Google Docs for the final report. An outline was created and members of the team were assigned portions of the report to write.

4/25 - Read logs

4/26 - Worked on the final report.


 * Plan:

Continue work on the final report.
 * Concerns:

Week Ending May 7, 2013

 * Task:

Review the current draft of the final report. Take into account comments made by other group members and make changes accordingly. Finish the report before the last class on Wednesday.


 * Results:

5/1 - Met in class. Discussed the current state of the final report and discussed changes and additions.

5/2 - Reviewed the final report.

5/3 - Read logs

5/4 - Worked on the final report.

5/7 - Made additions to the final report.


 * Plan:

Finalize the report. Make last minute changes before the last class.


 * Concerns: