About Test Files Number in qualitative_validation_for_net.ipynb in spite of Contrast Selection

[zaccharieramzi]

About Test Files Number in qualitative_validation_for_net.ipynb in spite of Contrast Selection

[zaccharieramzi]

The thing is that you are looking at only the test files with af=AF, so it's only about half of them, no matter AF. You have to handle the 2 different acceleration factors separately.

I actually already had answered this question here.

[fharman]

Hi Zaccharie, I was really confused and sorry for misunderstanding. When i encountered the same thing in commit bcd3fdd, i had to write down in github. But i still really surprised why i am choosing both AF=4 and AF=8. Because all test files are obtained with AF=4. Why do i choose AF=8 for the other half? Thank you for your return, Best, Zaccharie Ramzi <[email protected]>, 20 Kas 2020 Cum, 16:17 tarihinde şunu yazdı:

…

The thing is that you are looking at only the test files with af=AF, so it's only about half of them, no matter AF. You have to handle the 2 different acceleration factors separately. I actually already had answered this question here <#108 (comment)> . — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#109 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AO4IQ2R6NDEFHTZOYAFGPT3SQZT7RANCNFSM4T4YJEMQ> .

[zaccharieramzi]

No no, not all the test files are obtained with AF=4.
Half of them are acquired with af4 and the other half with af8.
That's why you have both in the public leaderboard: https://fastmri.org/leaderboards/.

[fharman]

Oww thank you very much.I don't know this. When i just looked at test datasets, i thought these must be AF=4.Because aliasing artifacts were not valid so much in inverse FFT2 reconstruction. Thank you again. Best, 20 Kas 2020 Cum 16:43 tarihinde Zaccharie Ramzi <[email protected]> şunu yazdı:

…

No no, not all the test files are obtained with AF=4. Half of them are acquired with af4 and the other half with af8. That's why you have both in the public leaderboard: https://fastmri.org/leaderboards/. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#109 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AO4IQ2UENCC4JBP623JMBM3SQZXBNANCNFSM4T4YJEMQ> .