Gtuner IV 1.07 PUBLIC RELEASE: Computer Vision

Official news and announcements from the ConsoleTuner team.

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby Mad » Fri Mar 06, 2020 9:55 pm

teckx wrote:just curious, didnt see it answered previously. if im playing on xbox does this require me to route it through my computer now for this scripting to work?

For CV scripts you need to live capture the game play either via the xbox app to PC or with a compatible capture card.
ConsoleTuner Support Team || ConsoleTuner Discord || InputSense Discord (2K / FPS)
Mad
Major General
Major General
 
Posts: 4532
Joined: Wed May 22, 2019 5:39 am

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby J2Kbr » Fri Mar 06, 2020 11:21 pm

USER101 wrote:I use Yolov3 in other projects sometimes and will be attempting to use model classification directly with gtuner. If anyone has any suggestions please fill me in on how you were able to get darknet working with Python 3.8. It works on earlier versions I just couldn't get it to work initially since gtuner is using the latest version of python. Here is an example of the open CV code I normally use. https://www.learnopencv.com/deep-learni ... -python-c/

According the PyYOLO documentation page:
After the installation the LIB_DARKNET environment variable should be set to shared library path. The path is required in runtime so my recommendation is adding this to your rc file. export LIB_DARKNET=<path_to_libdarknet.>

On windows, environment variable can be configured as the following:

https://docs.oracle.com/en/database/ora ... 7EC53807D0
ConsoleTuner Support Team
User avatar
J2Kbr
General of the Army
General of the Army
 
Posts: 20323
Joined: Tue Mar 18, 2014 1:39 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby USER101 » Sat Mar 07, 2020 3:31 pm

Ohh nice! Thanks @J2Kbr
User avatar
USER101
Sergeant Major
Sergeant Major
 
Posts: 100
Joined: Mon Dec 03, 2018 5:45 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby J2Kbr » Mon Mar 09, 2020 6:56 am

USER101 wrote:Ohh nice! Thanks @J2Kbr

You are welcome.

BTW, did you saw that the user Sucre published a GCV script based on YOLO Object Detection?

https://www.consoletuner.com/greslib/?w998

It seems something similar to what you are working on. Anyway, this is a confirmation YOLO can be used with Gtuner IV. :smile0517:
ConsoleTuner Support Team
User avatar
J2Kbr
General of the Army
General of the Army
 
Posts: 20323
Joined: Tue Mar 18, 2014 1:39 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby Tahsin » Mon Mar 09, 2020 12:56 pm

I thing detecting weapon by sond is more practical and no need a capture card or pc
but need a profetional to do this job :a1chill_angry1:
User avatar
Tahsin
Sergeant Major
Sergeant Major
 
Posts: 100
Joined: Wed May 15, 2019 7:57 am

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby antithesis » Mon Mar 09, 2020 2:07 pm

Unsurprisingly, I had a little trouble installing Python, so here's the idiot-proof instructions -

1. Download "Windows x86-64 executable installer" from https://www.python.org/ftp/python/3.8.2 ... -amd64.exe (v 3.8.2 is the latest version at the time of this post).
2. Open the downloaded "python-3.8.2-amd64.exe" file from your Downloads folder.
3. During installation, tick the box for "Add Python 3.8 to PATH" and select "Install Now". This is easily missed and is critical as it automatically adds the path to pip to the Command Prompt app.
4. Click search in Windows 10 (the magnifying glass icon) and type cmd. Open the Command Prompt app.
5. In Command Prompt, type (or copy and paste) "pip install opencv-python" without quotes.
6. In Command Prompt, type (or copy and paste) "pip install pywin32" without quotes.
7. Open Gtuner IV, go to Tools > Preferences > Computer Vision.
8. Click "Browse" in the Python Location field and browse to C: > Users > YOUR USERNAME > AppData > Local > Programs > Python > Python38 and click "Select Folder".
9. All 3 checkboxes in Computer Vision should now be ticked to indicate Python, OpenCV and PyWin32 have been configured correctly.

I tried installing to C:/Program Files/Python/Python38, however when pip installed OpenCV and PyWin32, the path pointed to AppData/Roaming/Python/Python38 instead and would not detect correctly within Gtuner IV. It also created permissions errors that were avoided by simply installing to the default Python path.

If we expect more than a dozen or two hardcore scripters to use Computer Vision, Python & OpenCV are going to need to be installed alongside Gtuner IV. I'm anticipating that'll happen, just flagging it as a potential issue for dummies...like me :oops:

Please add Elgato HD60 Pro to the compatibility list. I'll test Elgato Cam Link (1080p version) and Avermedia ExtremeCap U3 tomorrow. HD60 S is already confirmed (I think) but I can test it too.

Apologies USER101 for discouraging you from heading down this path a couple of years ago, but things change. We need an obvious advantage over the forthcoming C Zen (I have one, it's crap, but the public won't know that) and superior mouse & keyboard is one way to differentiate the old technology (CZ) from the new (T2).

CV has obvious benefits for mods like automatic, weapon-specific, anti-recoil based on the weapon model. Or simply mapping weapon recoil and packaging it into a Gamepack. Given the right people and tools, we can create some spectacular scripts using CV in Gtuner IV.

On the flipside, there is also the potential for aimbots, which we should discourage (my reason for cautioning USER101 against CV) as it could end up getting T2 banned (Sony and MS won't tolerate console aimbots). I'm not sure how that's going to be managed, but it's a risk that needs to be assessed.
Official Australian retailer for Titan One, Titan Two and XIM APEX at Mod Squad
User avatar
antithesis
Colonel
Colonel
 
Posts: 1912
Joined: Sat May 28, 2016 10:45 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby Baby Cham » Tue Mar 10, 2020 3:15 am

antithesis wrote:We need an obvious advantage over the forthcoming C Zen (I have one, it's crap, but the public won't know that) and superior mouse & keyboard is one way to differentiate the old technology (CZ) from the new (T2).


100% Agreed

antithesis wrote:On the flipside, there is also the potential for aimbots, which we should discourage (my reason for cautioning USER101 against CV) as it could end up getting T2 banned (Sony and MS won't tolerate console aimbots). I'm not sure how that's going to be managed, but it's a risk that needs to be assessed.


Respectfully, kinda disagree but that's just me. I do believe as long as we're not messing with the game code like what they do on PC's we'll be fine.
ConsoleTuner Support Team
User avatar
Baby Cham
Sergeant Major of the Army
Sergeant Major of the Army
 
Posts: 235
Joined: Thu Dec 15, 2011 3:39 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby J2Kbr » Tue Mar 10, 2020 10:36 am

Thank you antithesis for the step-by-step Python installation instructions. I am positive it will be of great help for new users interested to try the computer vision interface.

antithesis wrote:Please add Elgato HD60 Pro to the compatibility list. I'll test Elgato Cam Link (1080p version) and Avermedia ExtremeCap U3 tomorrow. HD60 S is already confirmed (I think) but I can test it too.

:smile0517:

antithesis wrote:On the flipside, there is also the potential for aimbots, which we should discourage (my reason for cautioning USER101 against CV) as it could end up getting T2 banned (Sony and MS won't tolerate console aimbots). I'm not sure how that's going to be managed, but it's a risk that needs to be assessed.

As you can imagine, for the past several weeks I've being experimenting with CV scripts and -at this point- I can ensure that a PC-like aimbot is not possible for several reasons, but mainly:

1 - detection: advanced object detection algorithms are very CPU demanding, such algorithms does provides real-time results;
2 - latency: my video capture card has 150ms latency (for streaming), it seems such latency is common on all capture cards;
3 - response: the challenge is not only detect a target, but also place the aim sights aligned with it. The aiming input on consoles is relative (not absolute) which would necessarily demand a procedure with feedback (move, check position, move again, check position, ... until aligned with target), which is not efficient and takes times.

In summary: yes, it is possible create an aimbot with CV, but its performance would be far worst than an average player.
ConsoleTuner Support Team
User avatar
J2Kbr
General of the Army
General of the Army
 
Posts: 20323
Joined: Tue Mar 18, 2014 1:39 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby teckx » Tue Mar 10, 2020 4:54 pm

this reads very similar to image search using a color hex.
User avatar
teckx
Sergeant Major of the Army
Sergeant Major of the Army
 
Posts: 194
Joined: Tue Mar 27, 2018 1:53 pm

Re: [UPD Mar-6] Gtuner IV 1.07RC (Computer Vision, Donation

Postby antithesis » Tue Mar 10, 2020 10:04 pm

J2Kbr wrote:Thank you antithesis for the step-by-step Python installation instructions. I am positive it will be of great help for new users interested to try the computer vision interface.

Happy to help and feel free to use it :)

J2Kbr wrote:In summary: yes, it is possible create an aimbot with CV, but its performance would be far worst than an average player.

Very good to know. I imagine this weighed into your decision to make CV public as a "good" aimbot could risk your livelihood.

I'm really looking forward to the mods that come from CV and making our own m/kb translators :smile0517:
Official Australian retailer for Titan One, Titan Two and XIM APEX at Mod Squad
User avatar
antithesis
Colonel
Colonel
 
Posts: 1912
Joined: Sat May 28, 2016 10:45 pm

PreviousNext

Return to News and Announcements

Who is online

Users browsing this forum: No registered users and 86 guests

cron