Reading and Researching

Caught up on a lot of the readings.

 I learned what dynamic time warping is.
I also learned contradictory things about sampling for sound.


Started coding the part for logrithmatic sampling, but decided a better spend of time may actually be to work on getting a good demo for next week. (My beta review is April 8th). So I've put aside the sampling code for now, and is working on getting the preceptor to understand their name.

After finishing that, i'll work on intergrating microsoft speech sdk into this whole thing.

If I'll have more time, then I'll go back and work on logrithmatic sampling. 

HCA Tree accomplished

The HCA tree is a weird form of huffman encoding.
The parent's weight is the average of the children weights.

Will look at how to sample over the weekend.

Before I sleep toda

Before I sleep today I will produce the HCA tree. I don't want to spend any more time being confused about which two nodes should be next to each other.

I'll be coding in Java, the algorithm will be similar to huffman encoding.

sudo code:
given confusion matrix, create nodes that represent each node's similarity. Use greedy approach (approximation) to construct the first pair, then proceed from there.

I don't particularly know if this'll work but I'll start from there.

Working on binary tree

I can't figure out how to construct it.
should 'p' and 'b' be closer or 'p' and 's'

The distance between the nodes is the distance between the values it represents. when things are in a binary tree, I don't really know how to keep them the same distance as the one from the paper.

looking at this link: http://ww4.aievolution.com/hbm1201/index.cfm?do=abs.viewAbs&abs=6470
It seems that there is a way of computing something called a "similiarity tree" from the similiarity matrix.

This study uses this paper: http://www.sciencedirect.com/science/article/pii/S0304397511007158
for constructing trees
 

Working in the lab

Set up Unity with the project.

At first it wouldn't compile. It took me until Thursday Morning to make the whole thing compile. I had to edit out all the GPU dependencies from the code.

Now it is friday morning. I just ran some propogations.

This is "t"

[0.01889388]:(onsonant_t@

onsonant_t:494.1513), (17_consonant_j_y.png:798.9461), (38_vowel_e_bird.png:800.9612), (07_consonant_tf_ch.png:918.0318), (08_consonant_h.png:932.5713), (37_dipthon_au_out.png:1202.313), (03_consonant_f.png:1400.457),....

this is the vowel from "bet"

[0.01889388]:(owel_3_bet@owel_

3_bet:139.6159), (26_dipthon_ei_eight.png:181.4589), (06_consonant_f_sh.png:189.3021), (33_vowel_u_foot.png:193.5665), (22_consonant_r.png:225.2484), (20_consonant_ng.png:296.1697), (16_consonant_d3_ju.png:301.0428), (15_consonant_3_jzu.png:310.3925), (03_consonant_f.png:318.9848), (23_consonant_w.png:427.7459),

 

 

Ah ha, I have a lot more to work on it seems. 

 

Oh, I didn't finish editing the paper. I don't think I'm going to until next week. I think it is more productive right now to keep on hacking at this code than writing the paper, especially because I'm extremely slow at papers.

Monday

Read through Pengfei's email Actually and took a look at the Unity code again.
I think this step will be more complicated than I originally thought.

Constructing the Hierarchy according to the paper "An Online Algorithm for Hierarchical Phoneme Classification". Sent Pengfei another question regarding how to connect this data, the matlab output and the Unix code together.

Will ask Mubassir about getting a computer in the lab. When I work on my laptop, the Unity code is way too slow. Unity crashed on me about 5 times today on my laptop.

my tree on paper right now

Back from spring break

Got off the plane a few hours ago.
This is my plan for this week:

Monday: Pengfei replied to my email about Unity. I will use the reply to take a look at the code base again. And put my data through the unity code.
Tuesday:  finish "Monday", and fix the proposal according to comments. (It takes me a long time to write)
Wednesday: Read some papers/research. From a battle plan for how to extract formants/other data. (matlab portion).
Thursday: Continue Wednesday. Start coding is possible.
Friday: catch up or more coding.

There's a Math midterm next Tuesday (the 19th). From Saturday until the midterm, I will not be very productive with this project.