/robowaifu/ - DIY Robot Wives

Advancing robotics to a point where anime catgrill meidos in tiny miniskirts are a reality.

LynxChan updated to 2.5.7, let me know whether there are any issues (admin at j dot w).


Reports of my death have been greatly overestimiste.

Still trying to get done with some IRL work, but should be able to update some stuff soon.

#WEALWAYSWIN

Max message length: 6144

Drag files to upload or
click here to select them

Maximum 5 files / Maximum size: 20.00 MB

Captcha
no cookies?
More

(used to delete files and postings)


Welcome to /robowaifu/, the exotic AI tavern where intrepid adventurers gather to swap loot & old war stories...


F = ma Robowaifu Technician 12/13/2020 (Sun) 04:24:19 No.7777 [Reply]
Alright mathematicians/physicians report in. Us Plebeians need your honest help to create robowaifus in beginner's terms. How do we make our robowaifus properly dance with us at the Royal Ball? >tl;dr Surely in the end it will be the laws of physic and not mere hyperbole that brings us all real robowaifus in the end. Moar maths kthx.
4 posts omitted.
>>7786 >but would certainly like a ballroom dancing bot. Seconded. What if /robowaifu/ were to focus explicitly and specifically on solving that single problem alone Anon? Wouldn't we in many practical ways be advancing the much bigger set of general motion solutions for robowaifus across the entire spectrum? Surely ballroom dancing represents a domain that really pushes the envelope of sophisticated, discriminating humanoid bodily motion, yet without going to real extremes like robowaifu gymnasts or qt3.14 robowaifu enforcers. Seems like a relatively balanced goal for us that is a real achievement in itself, but not too unreachable conceptually overall.
related xpost, I think. >>7825
>>7802 >...about on wheels? Because I reckon a robot skidding... That could be a good solution to start with. One of the requirements for dancing Viennese Waltz without getting too tired, is to maintain close contact and move past each other instead of trying to rotate. The cool thing is that the motion looks like continuous spinning when really you're stepping forward through your partner and doing a 180 degree turn and then repeat the action this time moving backwards. Do this cleanly enough and fast enough (6/8 timing music is FAST xD), then it will look like quite beautiful, the couple moving around on the floor. In practice, I rarely had a gal that I clicked well enough well enough with (and who cared as much about the dance as I did) to actually make Viennese waltz work. Basic form of this dance has the simplest steps, but one of the most difficult techniques. So mechanical legs are still probably necessary for such a fit (though may be wheels could be used where the feet would be?) >can ya tell I know nothing about ballroom dancing? :D I'll be honest, quite a niche hobby indeed XD >>7811 I like you thinking. I'm in a similar boat in that, I want a robo-waifu to entertain me (by communication and dancing). Other features are not necessary to me. So as you point out, developments should happen in parallel and in different areas. Perhaps much further down the line functionality could be combined (though I would not sacrifice ballroom abilihy for much else :P) Perhaps if we agreed on a common "personality core", which could be attached to different robo-waifu models, depending on the required task. Still get to have *your* waifu, just the outer hardware changes. >>7830

Message too long. Click here to view full text.

I'm using the videos of StatQuest (Josh Starmer) to learn about statistics and more: https://youtube.com/c/joshstarmer Here the basics: https://youtube.com/playlist?list=PLblh5JKOoLUK0FLuzwntyYI10UQFUhsY9
>>9065 I also want to recommend "Nudge Algebra" app for repeating some basic math skills, to prevent forgetting them, or for relearning if that already happened.

Work on my Elfdroid Sophie Robowaifu Enthusiast 08/18/2020 (Tue) 22:49:13 No.4787 [Reply] [Last]
Design and 3D printing is under currently underway to turn Sophie from an articulated doll into a proper robowaifu. I will post updates to design files on her Google Drive folder when I have confirmed that everything actually works smoothly. So far I've just got her eyes moving left and right, her lower jaw can open and close, and I am working on giving her neck two degrees of freedom. Links to file repositories below. http://www.mediafire.com/folder/nz5yjfckzzivz/Robowaifu_Resources https://drive.google.com/drive/folders/18MDF0BwI9tfg0EE4ELzV_8ogsR3MGyw1?usp=sharing https://mega.nz/folder/lj5iUYYY#Zz5z6eBy7zKDmIdpZP2zNA >=== -add file repo links
Edited last time by Chobitsu on 03/06/2021 (Sat) 00:01:43.
347 posts and 139 images omitted.
>>9101 LOL, the resemblance is spot-on!
Not exactly a 'related xlink' OP, but you might find this interesting if you follow the link. >>9080 BTW, you are just reaching the autosage bump limit. I'd suggest you create a new thread for your Elfdroid Sophie work. It's traditional to link back to the previous threads in a continuation thread, and to make a very obvious post at the tail end of the old ones screaming NEW THREAD. :^)
>>9160 I wondered when that would happen, thanks for the heads up! Oh and yes, that Glados robot is very impressive! Those cheap and clunky (but still pretty awesome) high-torque servos are found in everything from vending machines to electric windows and automatic doors, apparently. Which would explain why they are so mass-produced and readily available.
END OF ELFDROID DEV THREAD 1 I think. Of course other people could post here still but this is supposed to be the end of this thread LOL. SECOND SOPHIE DEV-THREAD BEGINS HERE: >>9216 Motor your way over there if you want to follow this eccentricity further!
>>9217 Good job Anon. I'd suggest you do the redirect more like this however: SECOND SOPHIE DEV-THREAD BEGINS HERE: SECOND SOPHIE DEV-THREAD BEGINS HERE: SECOND SOPHIE DEV-THREAD BEGINS HERE: >>9216 >>9216 >>9216 >>9216 >>9216 SECOND SOPHIE DEV-THREAD BEGINS HERE: SECOND SOPHIE DEV-THREAD BEGINS HERE: SECOND SOPHIE DEV-THREAD BEGINS HERE:

Message too long. Click here to view full text.


TalkToWaifu Robowaifu Technician 12/31/2020 (Thu) 19:53:18 No.7978 [Reply]
Anyone know what happened to the TalkToWaifu GPT-2 AI? I saw some screenshots of what I assume was the trained AI and it was great for a single dev. Recently went back to their Gitlab page and the user account had been deleted and remade, and the Github repo (https://github.com/kokubunji/TalkToWaifu) hasn't been touched in 9+ months. Anything out there more recent that this, an AI that has no human cultural restrictions? Corps love to add in shit to AI so they won't say anything racist, sexist, xenophobic, anti-Semitic etc etc you get the point.
2 posts omitted.
>>9114 MInor e-drama.
There's a slightly newer version of TalkToWaifu with some bug fixes here: https://gitlab.com/robowaifudev/TalkToGPT2 I've been working on an update that can use multiple different models, let chatbots talk to each other, and that has a training mode to guide the model towards generating better responses. TalkToWaifu was originally just some quickly hacked together code when GPT2 came out to show that it could be used as a chatbot. GPT2 doesn't make a good chatbot because it lacks an external memory to remember past events, and training a new GPT2 model from scratch would take a 100 years because of how inefficient it is. Fortunately there have been a lot of advances that make training a transformer model that performs nearly as well more feasible now, but more research needs to be done on getting it to work well with a recurrent neural network. Currently I'm experimenting with combining transformers with the non-saturating recurrent unit (NRU) that excels at learning long-term dependencies and using ReZero and parameter sharing to help the gradient reach deeper back in time. If it's successful the model will be capable of learning without training and only require a 2 GB GPU to fine-tune instead of a $9000 one. However, it will likely quickly forget what was learned until I figure out a better way to use neuromodulation to prevent forgetting. In the meantime there's someone else working on neural net chatbots here if you want to check them out. They're pretty good: https://github.com/JEF1056/Jade_T5 https://github.com/JEF1056/Jv6
>>9115 why the fuck is there E-Drama on a board that is designated for building robo-waifus. where is the connection?
>>9124 Thanks, mate.
Open file (17.34 KB 340x573 marry_me_pls_jade.png)
>>9124 >MARRY ME JADE!111 >no. Leld.

Papercraft waifu Robowaifu Technician 09/16/2019 (Mon) 06:21:35 No.271 [Reply]
Thoughts on making a paper waifu then adding robotics? I want animu grills but, most robots have uncanny 3DPD faces that aren't nearly as cute as a real waifu. With paper/screens, at least the face can keep the purity and beauty of 2D.
>>271 While I cannot help on the robowaifu side, there are moving paper toys from this thread back at /po/. They're cute and templates can be adjusted. They're easy to store, but be careful about paper getting dusty. https://boards.4channel.org/po/thread/574188
One of my favorite topics on /robowaifu/ actually. Not only is paper lightweight as a shell material, it's also easily 'editable' w/ no software required. It can serve as a base substrate to begin doing layup for other resin-based shell techniques. It can even serve as lightweight, easily-manufacture structural supports when rolled as tight tubes glued around collections of other 'tubes' like drinking straws. Combine these tensegrity-based structural frames on the inside, and easily-crafted shell designs on the outside, tie it all together with 3D-printed joints, connectors, bearings, pivots etc., and you have a lightweight robowaifu base from which to begin a design with. On top of all that it's cheap! I hope papercraft-waifu Anon finds us again and starts posting here again tbh.
>related xpost >>9176

HAPPY BIRTHDAY /ROBOWAIFU/ Robowaifu Technician Board owner 11/27/2019 (Wed) 17:27:05 No.1591 [Reply]
4th Birthday Edition greeting here: >>7224 Well it's Thanksgiving time again and that means it's /robowaifu/s birthday time too. It's been 3 full years since we began, and this last one has been a bit of a ride tbh. With the deplatforming of 8ch all of us who had communities there were mostly left out in the cold. Thankfully we've been given a new home here in our bunker at Julay, and a few of us have found our way back together here. Hopefully the rest of us will find Julay, and /robowaifu/ again as well. We've also made some new friends since we've been here, and I think we're in pretty good shape all things considered. On the topic of deplatforming, we have a new archival system in action that keeps a copy of the board up to date on my machine. Note that my original intent was to fully automate the archive.today archives, but it hasn't worked properly (they seem to be fighting against bots pretty hard now). I hope at least one other of you will also join in for the approach that's actually working, and use BUMP to keep at least one other copy of /robowaifu/ safe in addition to my own. You never know what might happen to my gear (or even to me), and it's just smart thinking to have at least two of us keeping this board safe. My intent is to finish up the system for board migrations (along with Robi), and then bring over other things from the old files and get this board in shape again. Then we can keep it fully backed up together. As you regulars know, /robowaifu/ is basically unique among imageboards and deserves preserving IMO. Any other takers for doing backups as well? I just use a cronjob and it's all fully automatic daily I never have to touch it. I'll explain everything in detail how to do it if anyone else wants to participate. One other thing; should anything ever happen to Julay, just regroup on Fatchan in our bunker-bunker there. Actually, I'm pretty upbeat about both the world of IB communities in general, and our own /robowaifu/ in particular. We all took a hard knock a few months ago, but we're back on our feet and in the fight again. Kudos to Julay and Smug in particular for their innovations with the webring. It's lent a resilience to the underlying systems for all our various communities that we've never had before afaict. This alone should help to ensure we have a long future of shitposting together anons. Anyway moving on, I hope this coming year will also see a resurgence in everyone's posts here about their personal projects. I intend to do so myself. It's always a pleasure to see anon's efforts and blog-posts about their current attempts and ideas, regardless whether they are successes or failures. Persistence is really the single most important factor for eventual success--even more than innate talent--and so just don't quit anon! Don't be afraid to get back in the game even if you met with failure before. Try something different this time, right? Industry-wise, I think that most predictions about the ever-increasing pace of technological innovations in the areas pertinent to robowaifu production are on track. I believe this will eventually also translate into much cheaper components and other resources that we all need for our projects. AI is gradually improving and becoming more accessible for the budding software engineers among us (and the old hands too). 3D printing is still improving and becoming more widespread and a bit cheaper in general. Mechanism designs are slowly becoming better suited to cheap production techniques. Electronics of all sorts keep getting better. Materials science breakthroughs also pop up occasionally.

Message too long. Click here to view full text.

Edited last time by Chobitsu on 11/26/2020 (Thu) 23:57:09.
42 posts and 19 images omitted.
>>8856 Thank, I'm going to try it. Thanks for the reminder.
Open file (38.37 KB 1315x527 ChanThreadWatch.png)
>>8856 I don't have a Raspberry Pi yet, but I did find this program called 'Chan Thread Watch' for Windows 10 that can download the html of each thread. Sorry if I'm doing it wrong - I'm a complete spastic when it comes to this sort of thing.
>>8860 Yes. I have played about installing a couple of different Linux distros on virtual machines in the past, and with the instructions about BUMP written on here, I should be able to make a proper backup. Thanks for pointing that out anon!
>>8900 Well if you have any questions just give a shout Anon. I've been doing stuff like VMs for about five years now (it's how I finally managed to escape the MicroStasi gulag).

AI + Brain/Computer Interface news & commentary Robowaifu Technician 09/15/2019 (Sun) 10:35:53 No.253 [Reply]
DARPA Wants Brain Implants That Record From 1 Million Neurons

spectrum.ieee.org/the-human-os/biomedical/devices/darpa-wants-brain-implants-that-record-from-1-million-neurons
18 posts and 6 images omitted.
>>5199 I guess what I'm thinking of would be a souped-up version of that PS4 VR game 'Summer Lesson', but maybe more Augmented Reality with IoT integrated so she appears to be cooking or cleaning when in fact it's the appliances running IRL and the animations are synced over the top of that. Sure, you'd still have to provide the correct ingredients for cooking in barcoded containers and go to pre-designated interaction points in your room/house but I think that would still be pretty awesome. Especially if you could take her outside with you on your mobile phone, too.
>>5200 >the appliances running IRL and the animations are synced over the top of that. I'm not too sure if the Visual Waifu got all the old archives synced across or not, but iirc this notion was discussed a bit there. Could be related, so might be of some interest. >>240
>>5185 >Dreaming is very nice, but getting things done is more important. I think we're all trying to create 'practical' things here Anon, even if they're virtual. Otherwise, we'd all be hanging out on /clang/ wherever that might be today, rip.
>>5199 >It's already better than reality if you ask me! /thrad

NLP General Robowaifu Technician 09/10/2019 (Tue) 05:56:12 No.77 [Reply]
AI Natural Language Processing general thread

>"Natural language processing is a field of computer science, artificial intelligence, and computational linguistics concerned with the interactions between computers and human (natural) languages."
en.wikipedia.org/wiki/Natural_language_processing
https://archive.is/OX9IF

>Computing Machinery and Intelligence
en.wikipedia.org/wiki/Computing_Machinery_and_Intelligence
https://archive.is/pUaq4

m.mind.oxfordjournals.org/content/LIX/236/433.full.pdf
31 posts and 10 images omitted.
>>8200 >using higher-order Markov chains for text generation Sure why not? >Would it be useful Yes, ofc. >or is it too simplistic? Again, yes. All NN-based approaches are too simplistic for now. GPT-3 is the best we've done atm, yet it's entirely brain-dead and doesn't 'understand' a word it's saying. It's behaviour is simply a yuge statistical model based on actual human's text output (well, mostly humans heh). Just start where you're at Anon. You'll get better as you go along. If you have an implementation of higher-order Markov chains for text generation that you can understand how it works well enough to modify it successfully to more closely align with your goals, then use it! You're already in the top 1/2 of 1 percent of the population in this regard. BTW, please share your progress and code here on /robowaifu/ as you go along, thanks.
>ivan-bilan/The-NLP-Pandect >A comprehensive reference for all topics related to Natural Language Processing >This pandect (πανδέκτης is Ancient Greek for encyclopedia) was created to help you find almost anything related to Natural Language Processing that is available online. https://github.com/ivan-bilan/The-NLP-Pandect
>>8496 Thanks Anon, repo cloned. That's a lot of links!
>>8496 Thanks I discovered EleutherAI through your post. Apparently it's a group attempting to create a fully open-source alternative to (((OpenAI)))'s GPT-3. We wish Godspeed to them, obviously. https://eleuther.ai/ https://github.com/EleutherAI
>related crosspost (>>10613, >>10614)

/robowaifu/meta-2: Electric Boogaloo Robowaifu Technician 05/13/2020 (Wed) 16:22:03 No.3108 [Reply] [Last]
Please Note: I'm going to be continuing with occasionally repopulating our board again with old posts. Feel free to post responses to them regardless, as I will generally be watching and can often respond appropriately to your posts. But be aware these will be 'log dumps' as it were from our original board. Specifically, these bump-bot re-posts will have rw@bump.bot in the name field (usually). Cheers. >--- This thread is for off-topic and general discussions. FYI, our bunker is at https://anon.cafe/robowaifu/catalog.html probably should bookmark that now Anon. :^) Also, let's discuss ways to get more people involved. How can we grow this board? As well, let's share general robowaifu ideas, etc., to help inspire each other. This thread is meant to improve /robowaifu/ in general way and be a place to hangout with loosely off-topic talk. previous /meta thread: >>38

Message too long. Click here to view full text.

Edited last time by Chobitsu on 10/29/2020 (Thu) 18:33:55.
349 posts and 107 images omitted.
Open file (1.96 MB 300x300 italian navy.gif)
You said there would be a new chatbot chobitsu.
>>8487 Heh, I guess you need to send moar pizzas first Anon. Also, grats you've just opened up for us a new subterranean zone. Probably filled with gold and jewels tbh. Any topic suggestions?
>>8487 When did he make such a promise and in what time frame? And when did we have a chatbot? GPT2 is a text generator. It doesn't know what it says, nor remember it.
NEW THREAD NEW THREAD NEW THREAD >>8492 NEW THREAD NEW THREAD NEW THREAD
>>8490 I actually wasn't quite too sure on that one myself heh. But I'm willing to give Anon the benefit of the doubt. We'll eventually get there with several effective chat systems I'm sure!

What can we buy today? Robowaifu Technician 09/11/2019 (Wed) 01:41:22 No.101 [Reply]
I've stated many times on this forum my skepticism of the "robo" part of "robowaifu". The reason why is because I've owned a number of dolls, including one full size TPE sex doll, and always been very satisfied with my purchases. The "robo" will be reality within our lifetimes but simply doesn't exist today. I'm starting this thread to discuss options that exist today which don't require advanced technology, products that can already relieve our loneliness. In particular I'd like to promote one cheap, freely available, and obvious option that you may have overlooked.

If you live alone and happen to have a few grand burning a hole in your pocket, you've probably already considered buying a sex doll. It's a good purchase for the money. If you haven't bought one yet, you probably either don't live alone, don't have that kind of money, or have a moral or practical objection to sex dolls. In that case you could buy a <$50 onahole-type or fleshlight-type sex toy, but that's not a "waifu" and isn't really more satisfying than masturbating with your hand. To alleviate loneliness, you need something that you can cuddle, than you can look at and smile, that you can care for, that makes you feel that you are not alone.

It may sound cringey, even pedophilic, but I want you to seriously considering buying a <$50 child's cloth doll. I've been cuddling and sleeping with an 80cm cloth doll for the last year and she has brought me much joy and comfort. A related option is a "dakimakura" (lonely pillow), but the advantage of the cloth doll is that it resembles a real human being, with a head and arms, and therefore is more fun to cuddle. My doll has made surviving this period of peak feminism so much easier, and honestly has probably saved me from the rope.

Here's the 100cm version of the doll that I bought. For cuddling purposes, bigger is better, so I recommend the largest doll you can find.
https://www.amazon.com/gp/product/B00ORWFLTW/

In this thread we can discuss sex dolls, cloth dolls, dakimakura, onaholes, and other options that exist on the market today.
25 posts and 16 images omitted.
I don't know at which point it started, but this threads contains a lot of postings which don't fit to the title. Somehow it got confused which a thread for homemade dolls and figurines, which is rather this one here: >>372 - the title picture looks similar. This one here is probably confused with the other one, and it's higher up in the catalog so it's easier to find, which is self-stabilizing. I only realized this some minutes ago, never really recognized the word "BUY" in the title.
>>8216 Good point. I'll happily give you mod privileges if you'd like, and you can edit to your heart's content Anon.
>>8219 Oh, one point on my todo-list I forgot about: Trying again to register at alogs. Last time Roddy didn't have much time and it coordination failed.
>>8222 OK sounds good. Just tell him I sent you if he asks, but he probably won't. I had let him know about you beforehand so yea.

Hold onto your papers Robowaifu Technician 09/16/2019 (Mon) 05:59:19 No.269 [Reply]
It's hard to keep track of all the developments that are happening in artificial intelligence and related areas. Perhaps we could share our favourite research papers to get a better feel for all the progress happening and what we need to do next to make robowaifus a reality.

I'm mostly focused on general intelligence but this thread can be for anything that has captured and held your attention for long periods of time, whether in speech synthesis, creativity, robotics, materials, or anything else.
Open file (26.90 KB 480x360 0.jpg)
Most may be already familiar with AlphaZero, the general reinforcement learning algorithm that mastered Chess, Shogi, and Go through self-play. It's a significant contribution everyone learning about AI needs to understand. It uses ResNets for a policy network and a Monte Carlo tree search to explore most likely possibilities and collapse the search tree into an improved probability distribution of expected outcomes for each possible move.

Original AlphaGo (January 2016)
<Paper:
storage.googleapis.com/deepmind-media/alphago/AlphaGoNaturePaper.pdf

AlphaGo Zero (October 2017)
<Publication:
deepmind.com/blog/alphago-zero-learning-scratch/
<Paper:
www.nature.com/articles/nature24270
<How it works (summary):
www.youtube.com/watch?v=MgowR4pq3e8
<How it works:
www.youtube.com/watch?v=XuzIqE2IshY

AlphaZero (December 2018)
<Publication:
deepmind.com/blog/alphazero-shedding-new-light-grand-games-chess-shogi-and-go/
<Early paper:
arxiv.org/abs/1712.01815
<Full paper:
science.sciencemag.org/cgi/content/full/362/6419/1140?ijkey=XGd77kI6W4rSc&keytype=ref&siteid=sci
<How it works (in-depth):
www.youtube.com/watch?v=_Z31-5D3RZg
<Tutorial:
web.stanford.edu/~surag/posts/alphazero.html

https://www.invidio.us/watch?v=2ciR6rA85tg
The papers are rather being posted in the relevant threads, but we could use this thread to share some sources for longer articles and papers to download. arXiv is a free distribution service and an open-access archive for 1,811,158 scholarly articles in the fields of physics, mathematics, computer science, quantitative biology, quantitative finance, statistics, electrical engineering and systems science, and economics. Materials on this site are not peer-reviewed by arXiv: https://arxiv.org/ "The Machine Intelligence Research Institute, formerly the Singularity Institute for Artificial Intelligence, is a non-profit research institute focused since 2005 on identifying and managing potential existential risks from artificial general intelligence. MIRI's work has focused on a friendly AI approach to system design and on predicting the rate of technology development." https://en.wikipedia.org/wiki/Machine_Intelligence_Research_Institute Artificial Intelligence @ MIRI We do foundational mathematical research to ensure smarter-than-human artificial intelligence has a positive impact. https://intelligence.org/ This posting >>7855 is based on an article from there. Looking into their research might be useful when our AI is getting more complex, not just be AIML chatbots. They have research, papers, and analyses on their website.
>>7857 >but we could use this thread to share some sources for longer articles and papers to download. Good idea, thanks Anon.

Report/Delete/Moderation Forms
Delete
Report

no cookies?