• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Toshiba CEATEC CELL demo

K

Kleegamefan

K. LEE GAIDEN
2005_ToshibaCell_Ref01.JPG


Toshiba brought 2 demos for this Cell devkit, well-known 48 MPEG2 streams playback and the new "Digital Kagami Type F" (digital mirror type F) demo. This new image-based rendering demo converts user's face into polygons and applies makeup to it by texture-mapping, template-matching and motion prediction by matrix calculation, and makes it into 720x480@30fps in realtime, on a half-mirror.

images787918.jpg


images787919.jpg.html


images787920.jpg


images787913.jpg


images787916.jpg


images787915.jpg


toshiba_cell01.JPG


toshiba_cell02.JPG


MOVIE








http://ascii24.com/news/i/topi/article/2005/10/04/658353-000.html
 
Wow, this deserves all kinds of OMGWTF attributes in the thread title, because that movie is pretty incredible.

I really don't understand how they are doing the face thing, but if it's as transparent as they made it seem, it's something that would turn many heads around if demonstrated with PS3 + Camera in some way.
 
Marconelly said:
Wow, this deserves all kinds of OMGWTF attributes in the thread title, because that movie is pretty incredible.

I really don't understand how they are doing the face thing, but if it's as transparent as they made it seem, it's something that would turn many heads around if demonstrated with PS3 + Camera in some way.
Click to expand...
well the cups demo at E3 was pretty damn cool, imo
 
Marconelly said:
Wow, this deserves all kinds of OMGWTF attributes in the thread title, because that movie is pretty incredible.

I really don't understand how they are doing the face thing, but if it's as transparent as they made it seem, it's something that would turn many heads around if demonstrated with PS3 + Camera in some way.
Click to expand...
Yeah, I'd like to hear more about how they did this. But I'm also guessing the SCC has a lot to do with the image processing. It must be more than a mere rasterizer/gpu. But the applications for something like this are incredible. If it's as versatile as it suggests, this could be used to create virtual wardrobes and hair salons where you can try out different looks before applying them. Hell, virtual tatoos would be priceless. No more buyer remorse. PEACE.
 
Doom_Bringer said:
what the hell is a super companion chip? btw that demo was radical!
Click to expand...
Toshiba's graphics/image processor that I believe was designed to work with Cell specifically. Looks like it used FlexIO, due to that bi-directional dataflow with Cell. Anyway, not sure if it's a 3D powerhouse or anything. There was a link to the patent on B3D a while ago. It's obviously quite capable...whatever the fuck it does. :lol

And yeah, someone else pointed out the speed of the update. Not a lot of delay at all. I would really like to know if this software (what's really doing most of the cool stuff here) is mostly utilizing Cell or the SCC. I'm curious to know if this is something that require special capabilities of the SCC, of it this could be implemented into a PS3 somehow. Eyedentify could actually become a worthwhile game at this rate. ;)

Oh yeah, there was also that Sony labs head demo that did something similar to this, but allowed free manipulation of the head/face once it was mapped in. Remember that from like a month ago? Japanese software devs are doing some really cool shit on the multimedia front, it seems. PEACE.
 
Pimpwerx said:
Toshiba's graphics/image processor that I believe was designed to work with Cell specifically. Looks like it used FlexIO, due to that bi-directional dataflow with Cell. Anyway, not sure if it's a 3D powerhouse or anything. There was a link to the patent on B3D a while ago. It's obviously quite capable...whatever the fuck it does. :lol

And yeah, someone else pointed out the speed of the update. Not a lot of delay at all. I would really like to know if this software (what's really doing most of the cool stuff here) is mostly utilizing Cell or the SCC. I'm curious to know if this is something that require special capabilities of the SCC, of it this could be implemented into a PS3 somehow. Eyedentify could actually become a worthwhile game at this rate. ;)

Oh yeah, there was also that Sony labs head demo that did something similar to this, but allowed free manipulation of the head/face once it was mapped in. Remember that from like a month ago? Japanese software devs are doing some really cool shit on the multimedia front, it seems. PEACE.
Click to expand...



Most likely, the SCC is derived from that cancelled CELL-based GPU....thats my guess anyway...

At any rate, imagine the business apps for this:

Opticians
Virtual Facelifts
Beauty Shops
Makeup Boutiques


Just off the top of my head...
 
mmm. Any idea how many cells it took to do that? Was it 1 cell plus SCC or another combination?

Man, if they can get this typa thing working with say the next gen eye toy.... can you guys imagine rendering your face in fight night next gen, tekken, SC.... mmmmmmmm, teh revolution indeed. ;p
 
The more I think about this ... damn, this could be a HUGE money maker.

If the CELL chip and SW solution can corner the market, they could sell a lot of units. All the trendy salons, tattoo parlors, and boutiques will definitely want to offer this tech to their customers. Shit, I'd like to use this the next time I go for a haircut.

This could actually be pretty big.
 
KLee,


BTW - I just made a purchase you might be interested in. I don't want to derail this thread, but in case you're unaware - Rocket 750's are being discontinued. Let's just say the price is hard to resist :)
 
This is incredibly cool! I canÂ’t wait to see the further applications that can be created with the Cell + Eye Toy for the Game Industry side of things!

A few years back I presented this exact idea to a friend of my father who owned his own barber school. Of course he didn't believe it would be possible (he was a stubborn elderly man) and I was adamant about how it wouldn't be difficult at all considering the technology and the advanced camera tracking software that has been developed. It would drastically improve several industries and speed up the process for all parties involved. Hell they could even further customize it for the costumer and allow them to provide their own personal designs.

The method I thought up of was using a series of points overlaid on top of the face (I'm guessing that was done with the brief overlay they placed) and using that to create an invisible polygonal model of the face on which to place the makeup/hair.
 
Fight for Freeform said:
Why does that newscast feature music from Super Monkey Ball?
Click to expand...

I wondered the same thing during the second tune featured.. haha

Anyway, I wonder how easy that was to do. Was it just a mater of sitting down and moving your head around so the computer could calcuate your face?

I would think it might be possible?
 
Okey.. it wouldnt be that hard to get a distorted face on the screen, but this seems to
scan their face somehow, to build a 3dmodel. I see it very difficult to get a depth in a flat screen. His nose could be like 40cm long, but this demo shows it just like his nose is...
 
AndreasNystrom said:
Okey.. it wouldnt be that hard to get a distorted face on the screen, but this seems to
scan their face somehow, to build a 3dmodel. I see it very difficult to get a depth in a flat screen. His nose could be like 40cm long, but this demo shows it just like his nose is...
Click to expand...

STI have done demonstrations with the CELL where 3d maps are rendered from 2d satellite imagery. This is probably an evolution of this technology. The aforementioned satellite imagery rendering supposedly requires a tremendous amount of horsepower.
 
"Height field data acquired by LIDAR (light RADAR basically) was combined with a color overlay taken via satellite imagery to create a realistic 3D rendering of the Mount St. Helens volcanic area."

This was even done on the Amiga in demos, using a picture with various colors where each color was a height above the reference plane.
This wasnt anything cool at all with the Cell demo.

The cool thing was that it was then raytraced in realtime, and that takes lots of cpu :)
 
Kleegamefan said:
Most likely, the SCC is derived from that cancelled CELL-based GPU....thats my guess anyway...

At any rate, imagine the business apps for this:

Opticians
Virtual Facelifts
Beauty Shops
Makeup Boutiques


Just off the top of my head...
Click to expand...
SCC is just an I/O bridge chip and has no graphics capability...
 
Why do I get the distressing feeling that the only gaming application this will ever see is some extremely disturbing personalized furry avatars for use in the next iteration of Final Fantasy Online?
 
You must log in or register to reply here.

Similar threads

Splinter Cell: Deathwatch Trailer
2
59
1K
EverydayBeast replied
SeverLink - Teaser trailer ( stealth game inspired by splinter cell and MGS )
18
123
Negotiator101 replied
Dragon Khan - Official Demo Trailer
3
210
Knight of the Round replied
Arm of God Demo Trailer
4
162
DaciaJC replied
Mole - Official Demo Launch Trailer
Trailer Indie 
2
93
roosnam1980 replied
Top Bottom