Final Assignment - Apply all your skills to solve a real problem!

Final Assignment - Apply all your skills to solve a real problem!

 

Final Assignment - Apply all your skills to solve a real

problem!

Due Thursday by 23:59 Points 30 Submitting a file upload

Start Assignment

For the final assignment, we prepared four competitions and one non-competition mode. The

description for the dataset can be found there: Competiton.pdf.

(https://myuni.adelaide.edu.au/courses/71454/files/10738541/download?download_frd=1)

If you choose competition mode, you can build up a team with a maximum of 2 students. Record your

teammate and the choice of your dataset through this link:

https://docs.google.com/spreadsheets/d/17yxAYl4uSRJaNEzxqpKZG7CPl0a1d0sSoAf2qXlT8RA

(https://docs.google.com/spreadsheets/d/17yxAYl4uSRJaNEzxqpKZG7CPl0a1d0sSoAf2qXlT8RA/edit#gid=

If you choose the non-competition mode, you need to finish this competition by yourself. Also, record

your information in the above link.

Please finish the pre-registration before the 3rd of May. After that, you will be automatically

allocated to one competition.

The guideline and detailed requirements for each dataset are available in the following notebooks.

The top three teams in each competition group will get bonus marks and a strong reference letter.

Facial: Facial_Expression_Recognition.ipynb

(https://myuni.adelaide.edu.au/courses/71454/files/10822045/download?download_frd=1)

Flower: Flower_Classification.ipynb

(https://myuni.adelaide.edu.au/courses/71454/files/10822047/download?download_frd=1)

Game: RockPaperScissor.ipynb

(https://myuni.adelaide.edu.au/courses/71454/files/10822051/download?download_frd=1)

Segmentation: Segmentation-1.ipynb

(https://myuni.adelaide.edu.au/courses/71454/files/10822943/download?download_frd=1)

Animal: Animal_Classification.ipynb

(https://myuni.adelaide.edu.au/courses/71454/files/10822044/download?download_frd=1)

2022/6/18 11:28 Final Assignment - Apply all your skills to solve a real problem!

 2/2

You should not modify the code of the evaluation metric (accuracy and FLOPs). You should not

modify the code for splitting the dataset.

Please submit your final solution in the notebook (convert to PDF) with your trained model and other

supported code files (in a zip).

Please submit your report in a separate file following the instructions.

Do not copy from each other. Because exactly the same answers will be very easy to find out after

ranking the accuracy.