PrioVR: Get your ideas moving(kickstarter.com)
kickstarter.com
PrioVR: Get your ideas moving
http://www.kickstarter.com/projects/yeitechnology/priovr-get-your-ideas-moving
9 comments
Has anyone tested the tech this is based on recently?
I tested it last year and it didn't work too well, but would be very interested to hear if they've ironed out the bugs since then.
I tested it last year and it didn't work too well, but would be very interested to hear if they've ironed out the bugs since then.
I've sent YEI a message to inform them about this thread, so hopefully we will have some official response.
Hello, my name is Steve and I am a developer at YEI.
We've worked a lot with the inertial sensing technology behind the suit, and have found it to be stable and well-suited to the task of motion capture. The biggest thing that can go wrong with this kind of technology is the presence of large magnets or ferrous objects nearby, but even in the roughest environments we've taken the current version of the suit to(mostly conference halls with giant steel beams running through them), we've always been able to calibrate it to work in the environment. There are also ways to automatically account for temporary magnetic distortions by relying on the gyroscope.
Outside of these sorts of distortions, inertial sensing has always seemed like a solid option for mocap to us, and one with a lot less potential for error than an optical or IK based system. I'd be glad to hear about any bugs anyone has experienced with inertial sensors in the past, and I can let you know what our system does to alleviate or sidestep those issues.
We've worked a lot with the inertial sensing technology behind the suit, and have found it to be stable and well-suited to the task of motion capture. The biggest thing that can go wrong with this kind of technology is the presence of large magnets or ferrous objects nearby, but even in the roughest environments we've taken the current version of the suit to(mostly conference halls with giant steel beams running through them), we've always been able to calibrate it to work in the environment. There are also ways to automatically account for temporary magnetic distortions by relying on the gyroscope.
Outside of these sorts of distortions, inertial sensing has always seemed like a solid option for mocap to us, and one with a lot less potential for error than an optical or IK based system. I'd be glad to hear about any bugs anyone has experienced with inertial sensors in the past, and I can let you know what our system does to alleviate or sidestep those issues.
VR seems like its getting incrementally closer to becoming a real thing. I wonder though, would that have been true without kickstarter?
I really hope this gets funded.
As popular as the Sixense Stem kickstarter is (exceeded its target within hours), its severely limited by the max number of sensors (5) in its working radius for full body immersion (joint positioning in particular) or multiple player usage.
We really need something like this in order to push the boundaries. The alternatives cost a small fortune.
(p.s. Not affiliated, just really want this to happen.)
As popular as the Sixense Stem kickstarter is (exceeded its target within hours), its severely limited by the max number of sensors (5) in its working radius for full body immersion (joint positioning in particular) or multiple player usage.
We really need something like this in order to push the boundaries. The alternatives cost a small fortune.
(p.s. Not affiliated, just really want this to happen.)
Well one alternative could be something like kinect, if you're willing to stay in one room. Although that might be harder to make really low latency?
The Kinect and other optical systems have a number of drawbacks we considered when deciding on inertial sensors.
Like you say, range and latency are issues, as well as occlusion, resolution of limb rotations, difficulties with multiple actors in one space, and general inaccuracies due to the system having to make a guess as to the actor's pose and position.
While the Kinect can work fine for simple games, we wanted the user's motions in game to feel natural and fluid, and to be able to extract the whole body pose even if part of the user's body is occluding another part(e.g. hands behind the back, crouched down).
Like you say, range and latency are issues, as well as occlusion, resolution of limb rotations, difficulties with multiple actors in one space, and general inaccuracies due to the system having to make a guess as to the actor's pose and position.
While the Kinect can work fine for simple games, we wanted the user's motions in game to feel natural and fluid, and to be able to extract the whole body pose even if part of the user's body is occluding another part(e.g. hands behind the back, crouched down).
Overlooking the optical limitations of the kinect, think of the potential uses for equipment like this.
Accurate motion capture anywhere (think of what we already achieve with GPS and limited device gyroscopic data alone), full body augmented reality (your body is the controller, lol) while walking down the street, the obvious VR and gaming implications, etc.
This is so much bigger than the ability to hit a ball around or wiggle your booty on a TV screen IMHO.
Accurate motion capture anywhere (think of what we already achieve with GPS and limited device gyroscopic data alone), full body augmented reality (your body is the controller, lol) while walking down the street, the obvious VR and gaming implications, etc.
This is so much bigger than the ability to hit a ball around or wiggle your booty on a TV screen IMHO.
Basically we need some good competition from the very start. Good for any tech.
Great, until you walk into a wall!
You could walk around the perimeter of the wall to 'calibrate' your safe environment, although thats dependent on the sensor drift (they say its very accurate).
I'm assuming with the combination of all the (multiple) inertial sensor data they could probably account for that to a pretty accurate degree.
I'm assuming with the combination of all the (multiple) inertial sensor data they could probably account for that to a pretty accurate degree.
Thats cool but still, what do you when you hit a wall in real life but could still go on in the direction in the virtual world ? That would need some confusing repositioning i would guess
I would expect the game itself to modify its virtual world to account for my physical boundaries.
360 degree treadmill: http://www.virtuix.com/about/
This does mean kicking is out of the question.
That and (realistic) jumping seem to be the main limitations of the omni at this stage.
I wonder if the omni harness was more like a baby bouncer than a fixed support if this would solve those issues:
http://static.toysrus.co.uk//medias/sys_master/8609611006271...
I wonder if the omni harness was more like a baby bouncer than a fixed support if this would solve those issues:
http://static.toysrus.co.uk//medias/sys_master/8609611006271...
Could have sensors warning you.
...breaking the immersion that you have tried so very hard to produce once or twice a minute so you can pause and reposition yourself in the real world room.
This seems like the biggest remaining problem for creating realistic VR games.
This seems like the biggest remaining problem for creating realistic VR games.
The end goal is of course to combine this system with a headset like the Rift and a treadmill like the Omni to offer fully immersive VR with boundless movement.
That being said, I still think PrioVR on its own has a lot to offer to games that hasn't been offered before. Imagine a fighting game where you dodge punches by actually ducking down, or a dancing game that can actually capture the full extent of your motion and respond accurately. Both of these concepts could be implemented without any fear of running into a wall, and should present a whole new way to interact with games.
That being said, I still think PrioVR on its own has a lot to offer to games that hasn't been offered before. Imagine a fighting game where you dodge punches by actually ducking down, or a dancing game that can actually capture the full extent of your motion and respond accurately. Both of these concepts could be implemented without any fear of running into a wall, and should present a whole new way to interact with games.
How is this an improvement over using a kinect to sense full body motion?
Kinect: is basically an optical sensor so it requires line of sight. It utilizes computer vision and a 3d depth map (generated by using an IR projector and receiver) to make an educated guess as to your bodies orientation.
Line of sight (needs to see you), Field of View (need to be within range), Body orientation (need to look humanoid, i.e. upright with 2 arms and legs), movement speed (not faster than the optical sensor or IR projector), clothing (not too baggy), lighting (low IR), etc, can all effect the (limited) accuracy of the kinect.
PrioVR: A system like this provides accurate positioning of all necessary joints in your body to form a true 3d representation of your skeletal orientation without any of the optical limitations or guesswork of the kinect. You could also use it outside, walking down the street, with a laptop in a backpack. i.e. it could be used for augmented reality in any situation.
In the future, I would expect a HMD (some kind of visor - an oculus with a camera in the short-term maybe), a full body suit (like this), and some gloves (the next project these guys will be working on) to be how we augment reality. I don't think we can let this opportunity slide, which is why i'm being so vocal about it!
PrioVR: A system like this provides accurate positioning of all necessary joints in your body to form a true 3d representation of your skeletal orientation without any of the optical limitations or guesswork of the kinect. You could also use it outside, walking down the street, with a laptop in a backpack. i.e. it could be used for augmented reality in any situation.
In the future, I would expect a HMD (some kind of visor - an oculus with a camera in the short-term maybe), a full body suit (like this), and some gloves (the next project these guys will be working on) to be how we augment reality. I don't think we can let this opportunity slide, which is why i'm being so vocal about it!
The new Kinect for the XboxOne seems to have vastly improved on the limitations of the Xbox360 Kinect. The PrioVR seems to be too ahead of its time. We don't have the battery tech to power all the tech you would be carrying on your body to play any immersive games for an extended period of time. Therefore, you would more than likely be tethered to a power source. If you need to stay in one location, might as well use a Kinect where you can be entirely free of wires. Powering just a visor would seem more feasible than a whole gaming rig. In addition, the current get-up seems cumbersome. I don't imagine that most consumers want to spend time strapping a multitude of sensors to their body just to play a quick game. The sensors themselves are great, but until we can shrink them down enough to embed them in actual clothing, I don't think we'll see them in commonplace usage.
I agree that the kinect is more convenient for casual gaming.
That said, I have yet to see a convincing demo of the new Kinect, so far only having seen increased resolution on the depth visualisation. By the very nature of the technology it will still have the same limitations as its predecessor.
The technology in the PrioVR may be a more cumbersome than ideal, but the proposed sensors are already smaller than those shown in the visualisation (http://i.imgur.com/oSBEloE.png - the middle ones are the new sensors). As for further miniaturisation, well it has to start somewhere (hence kickstarter, i guess).
As for power usage, this kind of device (low powered sensors with a single wireless hub) combined with a mobile phone and some kind of communication bridge would likely give you hours of full-system use (the PrioVR, this: http://www.kickstarter.com/projects/2041280918/vrase-the-sma... for display and some bridge to access the mocap data on the phone) off a simple 5Ah battery. Or you could go the oculus route, and just have a laptop in a backpack, powering everything through USB.
This isn't just about gaming. This is capturing full body movement for usage in a huge variety of applications. I'm worried that given the current traction this project has, it will fail to fund, and we're going to end up setting back immersive AR/VR for another decade.
That said, I have yet to see a convincing demo of the new Kinect, so far only having seen increased resolution on the depth visualisation. By the very nature of the technology it will still have the same limitations as its predecessor.
The technology in the PrioVR may be a more cumbersome than ideal, but the proposed sensors are already smaller than those shown in the visualisation (http://i.imgur.com/oSBEloE.png - the middle ones are the new sensors). As for further miniaturisation, well it has to start somewhere (hence kickstarter, i guess).
As for power usage, this kind of device (low powered sensors with a single wireless hub) combined with a mobile phone and some kind of communication bridge would likely give you hours of full-system use (the PrioVR, this: http://www.kickstarter.com/projects/2041280918/vrase-the-sma... for display and some bridge to access the mocap data on the phone) off a simple 5Ah battery. Or you could go the oculus route, and just have a laptop in a backpack, powering everything through USB.
This isn't just about gaming. This is capturing full body movement for usage in a huge variety of applications. I'm worried that given the current traction this project has, it will fail to fund, and we're going to end up setting back immersive AR/VR for another decade.
I agree that you have to start somewhere for miniaturisation, but if this technology is introduced prematurely and consumers don't take to it, you may be doing more damage than if you just let the tech mature in labs and commercial applications. Just look at the Virtual Boy. VR tech died at the consumer level when that failed.
I doubt that you could get hours of gameplay. Maybe if you were playing something very simple, you could do hours, but a fully immersive world would destroy battery life.
I doubt that you could get hours of gameplay. Maybe if you were playing something very simple, you could do hours, but a fully immersive world would destroy battery life.
I thought a quarter was pretty small (in UK here, so not sure of the comparison). In my opinion, this company is bringing what is currently a multi-thousand dollar setup down to consumer price levels. If they don't get the response they expect from this campaign, who is going to invest the millions to make it smaller? Surely that WILL be the death of these type of systems?
As for power usage, graphically intense 3d games on the iPhone 5 have about 2-3 hours battery life on a 1.5Ah battery (thats screen, CPU, GPU, etc), so I'd say its still perfectly realistic to expect hours. I don't know the power usage of this kit though, but i would expect it to be low solely based on its function.
As for power usage, graphically intense 3d games on the iPhone 5 have about 2-3 hours battery life on a 1.5Ah battery (thats screen, CPU, GPU, etc), so I'd say its still perfectly realistic to expect hours. I don't know the power usage of this kit though, but i would expect it to be low solely based on its function.
Has anyone done a Kinect + Oculus demo yet? Perhaps the last gen Kinect wasn't high enough resolution to get it right, but the next gen may be enough to do a somewhat convincing in-body effect.
Yes, there are a number of us in the oculus forums working on various solutions with (the current) kinect for positional data.
For me, oculus aside, facial capture is probably the most exciting thing the new kinect can offer (due to increased depth and optical resolution), but it still has all the same limitations as the last one for full body capture (as its basically the same mechanism).
There are a number of full body mocap solutions built by third parties that utilize the data from multiple kinects. This certainly helps, but it still can't be as accurate as something like this, solely due to you being a coloured point cloud blob instead of collection of bones in the softwares eyes.
For me, oculus aside, facial capture is probably the most exciting thing the new kinect can offer (due to increased depth and optical resolution), but it still has all the same limitations as the last one for full body capture (as its basically the same mechanism).
There are a number of full body mocap solutions built by third parties that utilize the data from multiple kinects. This certainly helps, but it still can't be as accurate as something like this, solely due to you being a coloured point cloud blob instead of collection of bones in the softwares eyes.
Hmm, very interesting. Can you give me any links for the facial capture stuff?
My full-body mocap needs are mostly sorted, but I'm extremely interested in facial mocap solutions right now because all the existing ones either exceed my budget or suck.
My full-body mocap needs are mostly sorted, but I'm extremely interested in facial mocap solutions right now because all the existing ones either exceed my budget or suck.
Faceshift.com is what I have played with most. It's amazingly accurate with even the current kinect, but starts at $800 for the indie version.
Brekel.com have a less expensive version, but I have no experience with it.
Open source alternatives seem thin on the ground, if not non existent.
Try faceshift, there's a free trial. It's mind blowingly awesome.
Brekel.com have a less expensive version, but I have no experience with it.
Open source alternatives seem thin on the ground, if not non existent.
Try faceshift, there's a free trial. It's mind blowingly awesome.
$800 for a facial mocap solution sounds, frankly, brilliant. I shall check it out on Monday/Tuesday.
How does the setup work? I'm assuming you have to spend a while tying expressions to morph targets on the model?
How does the setup work? I'm assuming you have to spend a while tying expressions to morph targets on the model?
Takes about 5 minutes to train the model using some simple expressions. Deals with detailed facial expressions, eyelid movement and iris tracking (the eye stuff is 'OK' - this is where the newer high resolution kinect should help, if supported).
How much more would it cost to get the wireless version that they show in the video and most of the pictures? This Kickstarter is only for the wired version which seems archaic in comparison.
Its important to note that only the sensors are wired. They connect into a hub on the chest (which transmits wirelessly), so you aren't tethered to anything.
The wireless sensor version (which they sell as a professional mocap device) costs about $4000 (which is still inexpensive for a current mocap suit) http://www.yeitechnology.com/productdisplay/3-space-mocap-st... (plus straps).
The wireless sensor version (which they sell as a professional mocap device) costs about $4000 (which is still inexpensive for a current mocap suit) http://www.yeitechnology.com/productdisplay/3-space-mocap-st... (plus straps).
Multiplayer fights are going to be bloody...
This is pretty awesome. I can't wait to play online multiplayer FPS games with this.
I'm so excited with all these recent advancements in consumer VR!