Closed jdcuevas-lbl closed 6 years ago
Not sure I understand the setting. Your model input is 64x64x64x8, correct? DeepExplain generates an attribution value for each input feature so your attribution map would have the same size of the input. Then it is up to you how to visualize such multidimensional vector.
Il giorno 23 lug 2018, alle ore 20:43, jdcuevas-lbl notifications@github.com<mailto:notifications@github.com> ha scritto:
If I wanted to generate a 3D saliency map of a 64x64x64 input with 8 channels, how could I go about doing it?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/marcoancona/DeepExplain/issues/16, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACLC5apzMpB5Gnfqbnw4qqvfzahRH3_xks5uJhlFgaJpZM4Vbf4S.
If I wanted to generate a 3D saliency map of a 64x64x64 input with 8 channels, how could I go about doing it?