## Jay :
1. Can u make a schedule for the future experiments? We need to finish it on december. It is including generate data from dmc3 dataset, and train the model using it.
2. Which dmc3 images will be converted to 1:5000 imaged? What is the criteria? For ex. Is that area has a lot of parcel or not.
3. How u gonna check the label?
4. Imo since we have a new dataset we may need to retrain existing model using new pineapple, rice, cabbage, and tea again-> oct (incase the result went south, we still have time to improve it)

## Hudson:
1. Schedule for the next half semester?
2. How many frames is needed to do classifications?
3. How many dmc3 images are needed to substitute those frames?
4. How u handle duplication when extracting the parcel from dmc3 camera images? (overlapping dmc3 images)
5. I found several frames from that exist from time to time, imo u can try to used it as real test. Check the ftp folder find the list. Please give me the list in excel-> check reference

## Perry :
1. Prepare a schedule for thesis

## Maysa :
1. Check the feasibility of duplicated frames experiment
2. make a flowchart for parcel change detection.
3. Check master student schedule, add missing necessary task

### :pencil:Notes
We probably need to check the correctness of the ground truth for the new dmc3 camera images.

### :book:Reference
[Duplicated frames](https://docs.google.com/spreadsheets/d/1DsenDeg8Yvi6HlKiwL4BmI-KufKymC0L/edit?usp=sharing&ouid=100769765831098120428&rtpof=true&sd=true)