The generated video doesn't follow the input image

[hitzhangjie]

The videos generated by this workflow do not follow the input images; they largely respond to the textual prompt instead. If we want to use this workflow to create short dramas, I need to ensure character consistency. Without good adherence to the image input, it is difficult to maintain consistent characters throughout the output video.
How can we adjust our workflow so that it follows the content of the input images more effectively?

[SyNgen-Smithe]

I've not tried them yet, but I saw workflows using these nodes along with the 10s nodes. They might be better for you.

[ulissescappato]

The videos generated by this workflow do not follow the input images; they largely respond to the textual prompt instead. If we want to use this workflow to create short dramas, I need to ensure character consistency. Without good adherence to the image input, it is difficult to maintain consistent characters throughout the output video. How can we adjust our workflow so that it follows the content of the input images more effectively?

Same here. Try disable "prompt_enhance" option.