Artist's commentary

Original

大败而归，垫草图跑ai还是会让画面太糊了 p02 我做的草图作画加上音效

最早开始做个人作品的时候，就有想法后期部分让ai帮我优化人体和上色了。这个14秒的出了后我看好多人想让我试试我也就找了个做ai还不错的朋友试了试。
大概花了500米左右，但成果看起来很拉。垫了草图视频后感觉ai还是太死抓视频里的人物形象了，导致画面bug非常多。
大败而归啊大败而归，看来我自己画动分，让ai帮我做后面完成品的想法还是太幻想了。
什么时候视频ai真的可以把抽象草稿识别好，并且还能生成出不会有抽象画面的精细人物，那以后web画师真就可以起飞了。可惜现在不行，我飞一半坠机了，man！

兄弟觉得也不太行感觉会被骂，所以就不和我搞共创了这期视频

p2是我画的那个草图动分，然后加上了ai视频生成的配乐和剪进去了原作剑圣怪叫的声音

The artist for some odd reason decided to use ai to do the polishing and coloring which ended up not working quite as well as they expected it to be. In their own words from the commentary (google translated):
"When I first started working on personal projects, I had the idea of using AI to handle the post-processing specifically for refining the character anatomy and coloring. After releasing that 14-second clip, a lot of people asked me to give it a try, so I teamed up with a friend who’s pretty skilled with AI to test it out.
It cost me around 500 yuan, but the results were terrible. Even after using my rough animation sketch as a base, the AI clung too rigidly to the character's appearance in the video, resulting in a ton of visual glitches.
A total flop, an absolute disaster. It seems my idea of drawing the rough animation myself and letting AI handle the final polish was just wishful thinking."

That ai ver can be seen on the source link, it's as bad as they described it as. And also idk if anyone even wants it but figured I'd also archive the ai version on aibooru. Why? Idk to point and laugh maybe