CSE 676: Deep Learning Deliverables Project 2 Solved

 CSE 676: Deep Learning   
Project 2 Deliverables 

Project 2 is for you to design and implement a project of your choosing. It is to be in the field of Deep Learning pertinent to the material covered in the lectures.

The deliverables have three parts: a report, a slide presentation and supplementary material.

1)    Report 

a)     It should have the structure of a paper, including

i)           Title and Abstract (one page)

ii)         Task Statement of Project (one page)

iii)        Method (Two pages or more) iv)       Results (one page)

b)    Include full references as footnotes where appropriate (Don’t include t at the end).

c)     Include figures (with reference as to where they came from)

d)    If it involved more than one person, indicate as to who was responsible for which part. If you don’t, both will receive one-half of the score for the project1

2)    Follow CVPR paper submission requirements and format for submitting the report, no other format will be accepted.

a)     Here is the LATEX kit: http://cvpr2021.thecvf.com/sites/default/files/2020-

09/cvpr2021AuthorKit_2.zip

b)    Report should be max 8 pages (excluding references)

3)    Should be submitted separately on UBBOX with filename as project2report_<ubitname.pdf (example: project2report_mshaikh2.pdf)

4)    Should be submitted on or before deadline

 

2) Presentation:

1.     It should consist of about 10 slides

2.     It should include four parts as in the report

3.     Use bullet points instead of full sentences

4.     We will have presentation for each group scheduled tentatively over 9th and 10th December

5.     Required to bring during the reserved presentation slot

6.     The presentation session will include Q&A and will be utilized for grading project 2 3) Supplementary Material: 

Required as it can help prove that your work and results are authentic.

1.     Can be utilized to show additional material which did not fit in the main report

2.     The .pdf in supplementary material has no length limitation

3.     Code that is submitted should be only essential files for understanding the implementation.

4.     No dataset should be submitted

5.     Appropriate weight file should be submitted

6.     Should be submitted separately on UBBOX with foldername as project2supplementary_<ubitname (example: project2supplementary_mshaikh2


Evaluation Rubric:

1.     Report (40%)

2.     Presentation and Q&A (50%)

3.     Supplementary material (10%) (may include videos, proofs, additional figures or tables, more detailed analysis of experiments presented in the paper, code)