[ReRemake重重制版][AI V2V]Iwara - Fuhua - Unholy Segs

コメント

这个视频就到此为止吧，我没有什么能优化的了。我一直在反复重制同一个视频，因为我发现在其他视频的AI V2V工作中遇到了很多新问题，我还在努力解决。我还有更多新的想法可以去测试，但AI视频的制作周期一点也不比原创视频简单，每一次测试都需要我付出至少两周时间的代价去验证。这么长的时间我原本轻轻松松的可以用来在Pixiv上发布超过十五个包含数百张AI图像的图片合集，但用在AI视频上，却只能换得一个“证伪”......我打算试一试最新推出的图像AI，比如MOSS和K3，他们都是写实类AI，可能和MMD动画不匹配，但在AI V2V这个领域，我已经穷途末路了，所有办法都值得试试......我删除了老版本的该视频AI下载链接，您可以在我的fansky主页获取这个AI视频的最新版本。它们是完全免费的。https://www.fansky.co/xiaoyingbi————————I think this video is done — there\'s nothing left for me to improve.I\'ve been remaking the same video over and over, because I ran into a lot of new problems while doing AI V2V work on other videos, and I\'m still working on fixing them.I have plenty of new ideas I want to test, but making AI V2V is not any easier than making original ones. Every single test costs me at least two weeks just to find out whether it works. In that same amount of time, I could have easily posted more than fifteen image sets on Pixiv, with hundreds of AI images in each set. But when I spend it on AI V2V, all I get back is a "nope, that didn\'t work"...I\'m going to try out the newest image AIs, like MOSS and K3. They\'re both realistic-style models, so they might not match MMD animation very well. But when it comes to AI V2V, I\'ve honestly run out of ideas — so every option is worth a try...I\'ve removed the download link for the old version of this video. You can get the latest version on my Fansky homepage — it\'s completely free!https://www.fansky.co/xiaoyingbi