- Jun 28, 2018
- 1,200
- 2,320
- Feb 1, 2018
- 348
- 561
Oh hey, what happened here?
Just to clarify things, the images are not dynamically generated. That would blow up the game to over 2gb with the AI models and renpy just barely supports python3. Instead, the images are generated during the game build process.
Yes, the graphics look bad. I've taken one of the state of the art AI networks and have been improving on it, but it's not there yet. On top of that, I've rendered my own dataset of nearly 40k images to help train the network. This is probably less than half of the number of images used to train networks like this, but new training techniques I've just adopted should help with that.
I expect things to get better over time. If you've followed the project since the beginning (almost a year now), you'll know it's come a long way and I'm extremely happy with the progress and the capabilities that the technology already has.
I appreciate the feedback on the story and for those brave enough to even try.
Just to clarify things, the images are not dynamically generated. That would blow up the game to over 2gb with the AI models and renpy just barely supports python3. Instead, the images are generated during the game build process.
Yes, the graphics look bad. I've taken one of the state of the art AI networks and have been improving on it, but it's not there yet. On top of that, I've rendered my own dataset of nearly 40k images to help train the network. This is probably less than half of the number of images used to train networks like this, but new training techniques I've just adopted should help with that.
I expect things to get better over time. If you've followed the project since the beginning (almost a year now), you'll know it's come a long way and I'm extremely happy with the progress and the capabilities that the technology already has.
I appreciate the feedback on the story and for those brave enough to even try.
Last edited:
- Dec 9, 2017
- 81
- 75
"The research and experiments to develop the initial models is a time-consuming process and uses intense compute resources (lots and lots of GPUs training for months at a time)."
Where the fuck are your GPUs from? Windows 95? Bruh, better graphics can be found in 30 year old games. Do Bethesda and EA both run your art design?
Honestly, that's the only quibble I have with this one. It's a fucking massive one, but it's the only one.
Where the fuck are your GPUs from? Windows 95? Bruh, better graphics can be found in 30 year old games. Do Bethesda and EA both run your art design?
Honestly, that's the only quibble I have with this one. It's a fucking massive one, but it's the only one.
- Feb 1, 2018
- 348
- 561
AI models are a bit different than a blender or Daz models. They take super computers to train (or a lot of time). Just the 0.3 release used over 15k teraflop hours of compute. In 1995 terms that'd be like 20 years of human time.
On the other hand, while it can take 1-3 minutes to render a high-def image in blender, a fully trained AI model can generate HD images at 2-3 per second. That's the promise of AI... if we can figure out how to get it to draw images better.
Last edited:
- Dec 16, 2017
- 882
- 1,925
I like what you're trying to do, I'll go through the game when I have time, best of luck to you o/
- Jan 22, 2018
- 3,224
- 3,935
This has to be an out of season April Fools joke! There's just no way any sane person would release this, let alone ask for support. I think even a man blind from birth can visualise something better than this nightmare inducing garbage fire.
You maybe trying to do something new and unique with AI, but this is unequivocally unpresentable and visually appalling.
Even chefs in training don't serve half cooked, bile inducing food. You should be ashamed of yourself for even posting this deepglugs.
PS- I hate being negative, but this just had to be said.
You maybe trying to do something new and unique with AI, but this is unequivocally unpresentable and visually appalling.
Even chefs in training don't serve half cooked, bile inducing food. You should be ashamed of yourself for even posting this deepglugs.
PS- I hate being negative, but this just had to be said.
I think trying something different is pretty cool and unique. Too bad most of these idiotic comments want another copycat daz game. I wonder if these people took to time to read what the dev said? "I cant fap to this reeeeeee"
- Feb 1, 2018
- 348
- 561
If you walk into the backroom kitchen, ignore all the signs that said "this might taste bad", and sampled one of the chef's experiments, you kinda get what you deserve. For my supporters it's more about the journey than the end-result of every build. They enjoy seeing the progress the AI artist is making with every new build and they enjoy being able to make their own custom builds according their tastes -a feature really only AI makes possible.
That said, what do you think of the story so far and the AI voices?
- Mar 11, 2018
- 491
- 707
stop complaining about the graphics and ask me a question there guys why the hell would i help a random guy to fuck the girls, if i have the powers i want them for me, this is the story of the game? help a crazy guy to fuck the girls? the google translator messes up the translation of the synopsis
- Feb 1, 2018
- 348
- 561
Euryale (the succubus) claims to get more lust energy from the subjects if they make love rather than just fuck. She has the power to seduce them all, but opts to try to get the "crazy guy" to fall in love with the others for the sake of more lust energy to gain..
- Feb 1, 2018
- 348
- 561
Not sure where the synopsis came from (let me know so I can fix it there too) in the OP, but the game is no longer a sandbox. It's just a typical VN now. Thanks!
- Mar 11, 2018
- 491
- 707
- Feb 1, 2018
- 348
- 561
The MC is a mysterious being that is helping Euryale for reasons Euryale doesn't quite understand. Can't give much else away without spoiling it. v0.4 will give more clues.and where does mc fit in that ?? or is he ''crazy''?
- Dec 9, 2017
- 81
- 75
Yeah, none of that makes it any better. Best to test things like this and improve them before releasing an inferior product. It's not really bragging if you're using awesome hardware/software to produce lesser content than the relatively inferior hardware/software can produce. I could potentially be a good football player but that potential is irrelevant in the real world. Potentially being able to create awesome things does not mean awesome things are created. All you did was confirm that it's better for content creators to stick with what actually works because advancements are not sufficient to justify the change.
- Feb 1, 2018
- 348
- 561
I appreciate your opinion. I'm not claiming this is a finished product. I'm releasing beta builds for my supporters and those brave enough to try and appreciate the progress. Feel free to ignore this and any future release.
Interesting analogy since no professional football player ever became professional except by earlier public exhibition... in front of other people... and often, they were at talent levels and capability far below that required of a professional player during their earlier exhibitions.
I am perfectly fine with this project failing in every other way than this. This is the baseline outcome. There are massive benefits in not only understanding what works, but what doesn't. How many times do developers fail in the same ways because the failures are not public, open and published? Perhaps its because people are afraid of comments like yours, they fail and disappear and no one ever learns from them.
I've already achieved beyond this baseline however. As one example, other developers have benefited from my usage of AI superscaling in this project from the
You must be registered to see the links
. Many of my supporters couldn't care less about the game and only care about the contributions I'm making to the AI and gaming industries.I guess what I'm trying to say is this: It's okay to not like the game, but please don't tell me to stop. I'm too stupid to give up before I've proven to myself that what I want is impossible to achieve.
Cheers!
- Feb 1, 2018
- 348
- 561
- Dec 9, 2017
- 81
- 75
It's a good analogy because potential without improvement is just that, potential. One can have potential all day but potential by itself means nothing. You clearly did not fully understand the analogy. I also didn't tell you to stop so don't put your misunderstandings on me. I meant exactly what I said so go back, read it again, and drop the bullshit passive aggressive stance. You're making improvements to the industry overall, cool, do you want a cookie for that? I'm gonna tell you like it is, it's not my fault if you don't like it. The visuals for your game suck. That's a fact not an opinion. I don't care if you like that or not and I'm clearly not the only one saying it. They will improve as the technology you're using improves. Cool. I didn't say I wouldn't keep an eye on it. If you're gonna tear at feedback just because you don't like the way it's presented then this isn't the forum for you and you're just gonna continue to get more and more salty as time advances. Actually, due to your whiny little attitude with your response to me, I could not give a flying fuck less about your game or your contributions to the industry. It's been doing just fine without you and it will continue to do so whether or not you passively fellate yourself over your work. Nobody is scared of honest feedback but most people definitely won't continue giving a toss about a dev who can't comprehend simple words. You clearly cannot handle brutally honest feedback appropriately and choose to make grievous misunderstandings the core of how you respond. I'm also not concerned with how little you'll be impacted by one person walking away. Respond to this or don't, I don't care, I won't be reading it anyway. Good bye!