hybridroulettecomputer.com V3 Development

[Steve]

Development of the Hybrid V3 roulette computer is progressing very well (nolinks.hybridroulettecomputer.com). The automated detection of the wheel in video, and automated prediction process is extremely complicated - especially since the angle of the video will vary, footage shakes, converting 60 frames per second video and ball position to precise timings etc.

One of the most challenging parts is for the computer to identify and track a particular wheel, and this part is now done. At nolinks.hybridroulettecomputer.com/v3.wmv is a video demo of the Hybrid V3 tracking a wheel in a video. It demonstrates:

   * How the wheel is selected and tracked: Importantly for the tracking, I selected a wheel in the far background and NOT the one right in front of the camera bearer (to make it much more difficult) and still you see it is easily tracked. The spins from the far back wheel can comfortably be predicted with a suitable video resolution.

   * How the computer knows if an object moves in front of the wheel: You can plainly see when the person moves in front of the wheel, the computer knows this and keeps the detection box in place. In such a situation, the camera-bearing player can easily step around the person as needed because the computer automatically learns what the wheel looks like from multiple angles. There is no need to aim the camera, and as you can see from the video, even a distant wheel which appears small in the video is easily tracked. This is a significant improvement on earlier models.

While this development is completed sooner than expected, there is still an enormous amount of work to be done. Now that the wheel position can be tracked, the computer needs to observe inside the wheel detection area to track the actual ball and rotor speed. This is actually easier than what is done in the video demo especially as we can use programming source code from earlier versions.

Nevertheless, the actual hardware requires a lot of development to make it small enough to be bodyworn. Overall, I expect full completion in 3-4 months.

[Steve]

Overall we are about 50% complete. We tweaked some of the reacquisition algorithms - this is where the wheel is not in the video (if the camera wearer turns away), then the hybrid needs to determine where it is again. A video of this component in action is at:

nolinks.hybridroulettecomputer.com/reacq.wmv

In this video, the left part shows the live video (in this case, just a video file). The top right shows what is being tracked. The bottom right shows the wheel being tracked in real-time.

During the video I click "Restart" to show reacquisition of the wheel. It is 100% reliable now. The end product will be able to track the wheel, ball and green zero even if the camera wearer does cartwheels, literally. The work achieved so far is military grade.

There is a lot more that is done that I wont show publicly. But I'm very happy with progress. Completion may be in roughly a month (although it is difficult to say for sure), but then we'll need to thoroughly test everything and tweak for improvements.

[Mr J]

Hey Steve, I'm not sure I have ever asked you this? Do you have (or did you ever have) a roulette method you played? Nothing with computers but an actual method. Just curious.

Ken

[Steve]

Do you mean like a "system" other than AP?

I played very frequently for about 15 years, now I mostly manage profit split partners. Much of it was "system" play, ie progression, outside bets etc and I had mixed results. Then I switched to AP. The learning and development has never and probably will never stop.

I dont think it can get better than wear a hidden camera and just be told where to bet - that's what the hybrid roulette computer is. If you can think of it, it can be done. I always tell new programmers if something cannot be done how I propose, I dont want to hear it cant be done - only other ways to do it.

With the "ap system" (ie genuinewinner.com), I'm about to release automated software that makes advanced ap as simple as "input data and get prediction". The missing link for my players is varying learning rates. Some just dont get it (the worse ever was bago although he was never a system player, and never understood the most basic things), and some are very quick to learn. On the note of bago, his arrogance screwed himself up and I came to learn his personal problems are rather severe. Anyway so the new automated software aims to turn an inexperienced player into a pro overnight. There is very little learning to be done. Regarding bets before ball release, I dont think it gets any better. Soon the user will input the data via a hidden mobile phone with cables, and they are told where to bet with a wireless earpiece or even phone vibrations.

Roulette has been a very long road for me. I dont need to do any of this anymore, but I love doing it. It is about time to let go though - probably another year or two for me, then my focus will solely be on energy research, with roulette more towards the hobby side. I already have enough revenue from businesses that require very little work, including profit split partners. I dont think I would never have a paersonal interest in roulette though and I will still maintain the forums.

[Mr J]

Okay, decent answer, thanks.

Ken

[Steve]

For the latest wheel tracking algorithm in action, see nolinks.hybridroulettecomputer.com/v3-afeb12.wmv

This video really gives it a good test with some fast, vigorous and unreasonable shaking of the camera. This algorithm learns the different angles of the wheel. If you watch carefully, you can see how well it learns from this short video - it learns both angle and scale changes. Especially angle changes are very difficult to do because with basic algorithms, the computer thinks the image has changed completely whereas we as humans can very easily recognize mere angle changes.  This is really not easy work to develop, so understandably we have taken longer than expected to get this far.

But as per my recent broadcast, the current developer is quite ill so we need to find a new developer to complete the hybrid v3 in a timely manner. To avoid further delays, the original developer will continue while the new one creates his own version. Then we'll just use whichever version is best.

[alio301]

 :lol:hahahahahahahaha. Your selling a computer that costs 3500 dollars. That apparently is accurate enough to tell you where the ball will land. Are you serious??? I know if i developed a comp. that could do this, i would just sit back and rake in the millions!! \I think we know why ur selling it....... 

[Steve]

the hybrid is $70,000 - $100,000 depending on options, the uber version is $30,000 for two player, and other versions are pay per 500 spins. most of my players with computers are profit split partners. yes the world is full of people who like act like they know something but know little. I think I live an ok life. nevertheless, thanks for your input

[Steve]

Below is a link to a video of the latest wheel tracking algorithm for the hybrid v3. It is important to note that the green box I select uses a tracking algorithm for the wheel's location, NOT the ball. So it is used to track where the wheel actually is, and other parts of the v3 software track rotor and ball movement. But to give the wheel tracking algorithm a real challenge, I selected just the ball in this video. The result is at nolinks.roulettesystemanalysis.com/m/balltrack.wmv

Despite it not even being designed for it, the software had no problem with ball detection and tracking - it really is military precision. When the technology is applied and optimized to track ball and rotor timings, establishing accurate timings very early in the spin (with faster ball speed) will be no problem at all.

While progress has taken much longer than originally anticipated (about 3 months behind), things are looking very, very good.

[Steve]

To make a few things clear, the main differences between the Hybrid (nolinks.hybridroulettecomputer.com) and my Uber computers are that the Hybrid:

* Takes timings automatically: no need to look at the wheel, unlike manual clocking errors.

* More accurate timings: With normal clocking computers, you click buttons to take timings (as the rotor and ball go around). The clicking errors are typically around 50ms. The Hybrid gets errors down to below 15ms, and it does it automatically - with your eyes closed, no clicking. I'm not talking about multiple revolutions, although this can be done too. I mean a single revolution. So it is far more accurate than any manual clocking computer. Unfortunately some people find it hard to accept this basic truth and vent on me personally. Its not my fault they only managed to copy 50 year old work. Some sellers claim their hardware is accurate to 0.0000001ms etc etc. Understandably, they are trying to sell their product, but manual clocking errors make hardware timing errors of 0.99999ms negligible.

To elaborate, my Uber (manual clocking) hardware has accuracy to 1ms. Ok so it's not 0.00001ms. But given it has 1ms accuracy when clocking errors are a whopping 50ms, and the Uber has far better algorithms than other computers, the end accuracy is far better. See nolinks://roulettecomputers.com/ddt.htm for example. See one of my devices that has 0.000001ms hardware accuracy: nolinks://nolinks.roulettesystemanalysis.com/m/micro.wmv .. but with prediction accuracy same as my the free roulette computer I offer, as explained at nolinks://roulettecomputers.com/algorithm.html and nolinks://roulettecomputers.com/comparson.htm

As for diamond targeting and what a computer needs to do, anyone only need to read all of the pages at nolinks://roulettecomputers.com/howtomake.html .. but if you expect ball deceleration rate will stay the same even with different conditions like ball track grit, and if you expect the ball to bounce the same way on different rotor speeds, well good luck

[Steve]

The progress video of the hybrid v3 is now available at Login

The timings from repeating the same video are attached.

This is from a standard PC and still the timings are almost all within 10ms accuracy. In fact they are mostly 5ms accuracy, but the few discrepancies are because of the computer's background processes. On the actual hybrid hardware, there wont be background processes that interfere. It is basically a PC stripped of background processes so that all it has to do is get accurate timings. Either way, the difference between 5-10 or even 15ms isnt significant. The typical timing errors from manual clocking computers is 50ms, so the hybrid is much more accurate, and completely automatic.

The progress video shows:

* The test interface

* Wheel location tracking (tracking where the wheel is when the camera is moving).

* Ball timings: I repeat the spin a few times so you can see the accuracy of timings, when the camera is moving more it would in typical conditions

Again these are the hardest parts done.

Players for this computer will be asked to arrange a meeting with me for training in about a month or so - just whenever the completed package is fully tested and ready to be used. I'll keep you updated.

[Steve]

Progress is extremely good and detections are 100% accurate in suitable conditions, and timings are comfortably accurate to within 10ms. So the main parts are perfect. Now most of the work is making it EASIER to set up. Already it is very simple but there is still room for improvement.

The latest video update is at nolinks.roulettesystemanalysis.com/m/v3ir.wmv
This time it is WMV format and youtube.

nolinks://nolinks.youtube.com/watch?v=ObRgpOd0110

About the Video:

* The wheel's location is tracked, and the ball is searched for in the box.

* Another part of the software shows the extracted video for analysis. I created a "picture in picture" video so you can see what the software sees.

* The ball is marked by an orange box

* Because video frame rates would normally have significant errors, the position of the ball relative to a fixed reference point is accounted for. So if the ball is say 15 degrees off, the software determines the precise timing based on the angle discrepancy. It is accurate to within 10ms, which is 50 times more accurate than manually clicking a button as normal roulette computers do.

* Development has cost me now nearly 3 times as much, and many times longer than originally anticipated. But because of the time to develop, purchasers will only pay the mid-range of original estimates.

* It is difficult to give a time estimate for release. First we have to complete making setup easier, then check for any bugs. Then a few of the other minor components need to be completed. I will give updates when I know an accurate estimate.

[Steve]

Below is another demo of the hybrid v3. This version uses mostly 1 detection per pass. The latest version can detect the ball 5+ times on one revolution (pass), so it is much more accurate although I havent shown this version because the interface has parts that arent for public viewing. The latest version gets accuracy down to about 1ms, for a single pass of the ball. Normal computers can only do about 50ms. Consider the difference in capabilities, especially how quickly it can get accurate prediction, and the algorithms it uses for calculations, and it isnt much of a comparison to computers of the 1980s.

Everything works the same way for live footage except the frame rate is better. To use the hybrid, when the spin begins, the computer either detects the spin has begun or the user clicks a button which tells the computer to start tracking. The end result is the predictions can be sent completely covertly to any number of players around the table.

To stay updated, please subscribe to my blog but please dont contact me about it until I announce the invitations, because people are already asking about it.

nolinks://nolinks.youtube.com/v/VOPMJmp7nw8?rel=0

More about the history of these devices is at nolinks.hybridroulettecomputer.com and this is the 3rd version of them.

[krist]

to buy this computer I have to submit $1000 deposit, which is non-refundable, don't I? Isn't it too much?

[Steve]

No if I dont show you what I agree, I refund it. Anyone serious about it needs to speak to me and can easily organize any proof they want.

[Steve]

-