Group 4

01/28 Meeting Report

We do final rehearsal!!

I hope we will do our best!

01/14 Meeting Report

We practiced during our meeting and evaluated each other's presentations. We found that our presentation exceeded the allotted time provided by the organizers. Hence, we tried exchanging suggestions on how we could minimize the discussion on certain slides of the presentation. We also decided to conduct our final rehearsal next Saturday.

12/27 Meeting Report

JAPAN TEAM

- Already finished with their model but are having difficulty with developing their application.

PHILIPPINES TEAM

- Already finished with model building with 96% - 97% accuracy.

- We will update our manuscript and find more pictures.

- We will polish our algorithm based on the comments of our research adviser.

For the report on January 28:

- We decided on the flow of presentation this December 17.

- Both teams have decided that on December 20, we should have a presentation template.

- On January 13, we will have a draft #1 of our presentation and our teachers would comment on it.

- On January 14, we will have our first practice of our presentation.

- And then we have the remaining two weeks to practice for January 28.

12/03 Meeting Report

our program will finish by December

On January, we will make presentation

11/05 Presentation Report

PHILIPPINES TEAM

- Data had already been loaded to TensorFlow.

- The renaming of the data (1000+ pictures) is already in progress.

- A small data set is being temporarily used from the data we gathered to make the program work, which will be replaced once all the images in the Google Drive folder have been properly renamed and organized.

- A number of parameters have been set in the code for the resizing and rescaling of the pictures. For example, for flipping, the total images will be increased to a factor of two.

- More parameters will be added in order to increase the number of the data.

- For next week, the Philippines Team will update their progress on data augmentation and data renaming.

JAPAN TEAM:

- Had updated how well their research presentation went in Kyoto

- A friend had sent them more pictures of Taro infested with Leaf Mold.

- They will try to collect more pictures of Leaf Mold and if all things go well, they will code for a program that will differentiate healthy Taro, Taro Leaf Blight, and Taro Leaf Mold.

- Their data augmentation is done already.

- For next week, the Japan team will update their progress on resizing the pictures they will be using for their coding.

10/08 Meeting Report

- For today's meeting, both teams updated our progress on our research.

-> Japan Team had collected more pictures and had stated how many pictures are needed, with each type of picture having at least 1-100 pictures.

-> A program where a drone would be used will not be needed.

-> A PC can be used to develop a code for an android. An emulator should be used.

-> The Philippines Team had collected more than 200 pictures of healthy Taro and Taro with diseases.

-> A research adviser had also been consulted, these are the things he had said: app development would be difficult to finish in 3 months, recommended using OpenCV and anaconda in programming, minimum of 100 pictures can be used in the first training, recommended using AndroidStudios for app development, early Batch of Training: 50% of total data should be unhealthy; 50% should be leaves with Taro Leaf Blight, later batches of training - 75% of the total data should be healthy; 25% will be Taro leaf Blight, suggested to edit the scope and limitation of our study, gave other related studies to help us in our methodology

-> Both groups will not be meeting for next week as Japan team would be busy.

-> Both groups also had agreed that the almost finished algorithm should be done by the end of October or start of November.

09/24 Meeting Report

126days to ICRF

Our Progress

PSHS Bicol campus

・We are collecting photo which is taro disease.

・We find teacher who teaches programing.

Ryugasaki 1st high

・we found paper about reaf mold and mosaic vius.

・Our teacher found Saitama rink. So,we will send an E-mail.

・Also,we have to contact these paper's Research Institutes.

・We are making programing which photos taken need to be moved to the Google Drive.

・We are learning about App’s layout and Event handler.

・We are struggling to understand java.

Both of us should collect photos.

09/17 Meeting Report

We discussed the tasks each team has accomplished. The Japan team is currently working on the program and still needs around 20 pictures to successfully run their program. They are also currently studying how to use Android studios. The Philippine team is also currently working on the program. We have made a prototype as a guide in making the actual program using the images gathered by the team. We are currently having a hard time finding pictures of leaf mold and mosaic virus. The first solution we agreed on is to further search the internet for more papers regarding the latter diseases. The last option is to just settle with Taro Leaf Blight as the only disease to identify using the program to be made. We have also changed the schedule of the meeting from 5 PM JST every Saturday to Saturday, 8 PM JST for everyone's convenience. Our task for this upcoming week is to look for more papers and continue learning about machine learning and building the program.

09/05 Meeting Report

We do our meeting for the first time in two weeks.

We shared our photos.

Also,we shared these photos at google drive.

PSHS wii add their photo after meeting.

Our tasks are to corect mosaic and to make programing.

Ryugasaki 1st installed emulator.

FSHS starts school.

So,Our meeting will be Saturday at 17:00@zoom.

08/15 Meeting Report

First, we reported our progress from August 1 up to last week. Japan team has collected a total of 72 images of leaves with Taro Leaf Blight. They have also contacted prefectures for data collection. At the same time, they have been developing the program and started learning about app development. As for the BRC team, we have contacted universities and research centers and are still waiting for their response. A professor of one university contacted us and said that she might provide us with data but recommended that we visit the farm ourselves to gather more data. We have also been learning how to build a program. The Japan team shared with us the workspace for the program development where both teams could work simultaneously. Additionally, we established a google drive folder where we would place the gathered images. Both teams have decided to gather more pictures of Taro and will proceed with the program development while waiting for the responses of the professors, universities, and prefectures we contacted.

08/01 Meeting Report

We shared how will we collect taro disease images on this meeting. Both schools are sending messages to university professors and research institutions in an attempt to obtain images. Programming and application development are also advancing at the same time.

Our next meeting is the week after next. So, we have to do what we can. For example, collecting images, learning programming, etc.

07/25 Presentation Report

- The students from Ryugasaki Daiichi shared the responses of the people and prefectures they've contacted regarding the collection of diseased Taro pictures. They've also shared their attempt to code the program.

- The students from PSHS-BRC shared the detailed steps to code and how to turn it into an application.

- We talked if we will make an application.

- We discussed about how we will solution our problem regarding the difficulty in collecting pictures.

- By next meeting we would update each other of our progress in collecting pictures and contacting various centers, farms, prefectures, and etc.

07/18 Meeting Report

・We shared information about taro diseases in each other's countries.

・We discussed creating a taro disease app.

・We discussed our upcoming schedule.

07/11 Meeting Report

• In the past week, both groups from each country had a meeting with their respective sources to gather more information about the Taro production in their areas. Today, we each did a presentation on the information we have gathered.

• By the end of the meeting, we have decided to do research on different lists of diseases according to what is most common in our countries.

• For next week, we will create a timeline of the activities and tasks we will do for the following months. This includes how long we will work on this task and so forth.

• In addition to this, we will also do further research on our respective list of diseases.

07/04 Meeting Report

• We shared information about diseases of Taro.

• By the next meeting, we will go to farm and get information about disease.

06/27 Meeting Report

• We agreed on our final topic.

• We have decided that Taro will be the crop we will focus on for our research.

• We agreed that for now, we will be using Jupyter Notebook as the platform for programming, but could be subject to change if we encounter future problems.

• By next week, we will list and discuss the differences between the Taro crop's growing methods, the plant's diseases, and how different the diseases are from each other, in Japan and the Philippines. We will also be discussing if we should consider developing the program into an app.

06/20 Meeting Report

We talked about taro.

PSHS want to find a disease to use programing.

We want to find it to use smart phone.

Ryu1 said "we want to solve water purification".

06/13 Meeting Report

For today, we introduced ourselves in order to know more about each other. We also introduced our schools and home country. Afterwards, we decided on the day that we will meet weekly, as well as our social media platform where we can communicate with each other.