Flows become strange when set n_shot >= 1

[HrsPythonix]

I train the few-vid2vid with FaceForensics dataset, and set n_shot to 2 or 4 to try to use multiple reference image during training, but the flows become dissimilar to the flow result when n_shot is default (= 1). Can someone explain if this is a bug? Thanks in advance for your answers!

Flows of n_shot=1:

Warping result:

Flows of n_shot=2:

Warping result:

[YixiangMao]

Hi, I went into the same issue. I was wondering have you figured it out and would you mind sharing any experience? Thanks a lot!