# 12/26/24 Meeting Notes #16
# This Week's Progress 這週進度
## Classification Model #1
Latest dataset count:
- 01Dress = 3778
- 02Shirt = 4150
- 03Tshirt = 4405
- 04Hoodie = 1555
- 05Sweater = 2066
- 06Polo Shirt = 1404
- 07Pants = 5145
- 08Shorts = 4027
- 09Skirt = 3079
- 10Flats = 3896
- 11Heels = 3898
- 12Shoes = 1776
- 13Sneakers = 3909

Traning accuracy = 89.14%
Testing accuracy = 89.6%

Confusion Matrix

## Classification Model #2

Latest dataset count (training):
- Bohemian = 227
- Casual = 1588
- Formal = 1528
- Semi-Formal = 1532
- Sporty = 349
- Streetwear = 380

Testing dataset:
- Need to run *Classification Model 1* script to the VISELLE dataset to classify them into tops and bottoms
- A python script needed to make outif combinations
Output (not VISELLE dataset):

## Generative Model
- Found masking module which is not given from the paper, but still haven't installed
`找到了一個masking的模型，還沒下載`
[DenseCLIP](https://github.com/raoyongming/DenseCLIP?tab=readme-ov-file)
- Installed environment for landmark module but haven't tested
`landmark模型的環境已準備好(tensorflow 2)`
## NSTC Project

did the literature review for classification model 1

Proposal Draft (https://docs.google.com/document/d/1BVFdVj92j1GGW5aQRl5k04Tw_gCovDSvi79kTu5ItPA/edit?tab=t.0)

# To Do 需做

# Next Weeks's Meeting 下週會議

---
Previous:
Next:
Full Content List [here](https://hackmd.io/@emps-113up/full-list)