Wan POV

Version 1.2 Update: Higher quality and expanded dataset. It can now do where the woman keeps and you can see the dripping out of her mouth. The I2V version can handle images where a does not start out in the image.

Note that all of the previews were created using the lightx2v (self forcing) lora. You may get better movements without it, but I didn't do much testing without it. The workflows I used can be downloaded by getting the "Training Data".

T2V tips: It actually seems to still maintain motion very well at 0.7 strength, and you'll get more facial variety that way. I seem to get best results at 0.95 strength for whatever reason. You can add in a lora to help it out.

I2V tips: If the does not exist in the starting image, you probably want to add a lora to your workflow so it has a better shape. You can trigger it with something like:

A  appears from the bottom of the frame centered at the bottom and pointed straight up.

If you want to appear:

White  shoots out of the .

This was trained on 4 different angles, here is how to trigger them:

A woman is lying on her stomach between the legs of the viewer and performing  on a man. Her head moves up and down as she sucks the .

An overhead view of a woman kneeling between the legs of the viewer and performing  on a man. She moves her head back and forth as she sucks the .

A woman is leaning over a man positioned in between the legs of the viewer and performing  on a man. Her head moves up and down as she sucks the .

A woman is kneeling in between the legs of the viewer and performing  on a man. Her head moves up and down as she sucks the .