Hey, thank you so much for your comment! It really means a lot to me!!! I started a new job a few months ago and I don't have too much free time at the moment, but I will make the Vertex AI tutorial as soon as I can!!
Hello ,author, your work is excellent, I would like to ask, when you do the labeling work in Makesenseai after taking screenshots to obtain data, do you need to label the fruit that has been cut?
Muito bom o tutorial Moises. Curti também o video anterior! To na missão de fazer um "jogador automatico" pra buscar item pra mim no Ragnarok, e o seus videos estão me ajudando muito!
Fiz o modelo no tutorial do seu video passado mas usei poucas imagens e ficou inconsistente. tem Como eu adicionar mais imagens ao modelo atual ou deveria fazer outro do zero?
Hey, I like the subject, I havera great video suggestion for you, make a new video about How to connect a Yolov8 Custome trained model to Connect toAny Game Window. In your first video you have been using Yolov4 , the current version is V8 which has so many options, for Detection ,Tracking, Segmentation and so on, now it is even easier deploying implementing for custom projects.
Hi there, thanks for your suggestion! I used v4 because I had used it in the past, but when I have the time, I intend to do an updated version with v8!
Make tutorial for character that use autopath, example in diablo 2. How can i make my character follow a specific long path. and walk in that specific long path only from location-A to location-B. Thank you! Pls make a guide 🙏
Hello! It will be hard to explain this in a youtube comment, but basically, in the variable "new_coordinates" you have a list with all the detected objects, each object being an dictionary with keys x, y, h, w and class_name. basically, you need to use the x, y, h and w to get the coordinate and move the mouse to the first object, the center of the first detected object will be: (new_coordinates[0][x] + new_coordinates[0][w]/2, new_coordinates[0][y] + new_coordinates[0][h]/2)
Hi sorry i'm kinda new to this, so i don't really understand everything. I would like to do the same with a 3D combat game, but by making the agent attack and dodge. Can i do this with that technology? or can i mix YOLO and Pynput in order to do that ? Thanks a lot
Hi there! Well, you can do that using YOLO and Pynput, but it will be very challenging. In my case, I only had to detect fruits, but in your case, you will need to not only identify your opponent, but when your opponent is going to attack you so you can dodge. For your purpose, if you have the opportunity to try a packet sniffer (a program that intercepts the communication between a client and a server) to detect your opponents' actions, instead of using an image detector, I highly recommend starting there.
hi I still use Bluestacks to test different games, but the mouse pointer doesn't point to Bluestacks itself, but to the detected window. What could be the possible reason?
Hello, I recommend you move the Bluestacks window to the top left corner of your screen. The mouse movement considers this position as the starting point. Could you try it and let me know if it worked?
Sure! basically, in my video i moved the mouse when i detected a fruit, and in your case you must move the mouse (and maybe press the left mouse button to shoot) when you detect a specific monster in re7
Thanks, man!! I will make more videos when I have the time. I believe that my next videos will be about how you can use object detection without machine learning because for tasks like detecting objects with simple 2D graphics (such as pixel art), there are a lot of other ways to do that using OpenCV.
hi, the window "DETECTED OBJECTS" renders a blackscreen, i have tried to launch bluestacks in admin but it doesn t change anything. Do you know what can i do?
Hi there. I had the opportunity to try to help one person with the same problem in a Discord call, and I believe that this is related to some anti-cheat of your game. Because in this Discord call, the code worked in some games, but in others it generated black screens (I also suspect that games within Steam may generate black screens as well). I suggest you try with other games to see if that is your case. For me, it worked with the games Terraria, Ragnarok Online, Diablo 2, and Fruit Ninja using the BlueStacks emulator. I am trully sorry I can't help any further, but I didn't encounter this problem while creating the tutorial.
Brother great explanation in simplest way loved you videos, but can you tell me what code to write if i want the crosshair in game to go to the custom trained object"s center. Hope to hear from you soon, waiting for your future videos.❤❤❤❤
Hello! I don't know if I understand correctly, but if you want to get the center of the detected object (x, y), you can just get the return of the function improc.process_image(ss) (that will return a list with the detected objects), and for each one, you can get the x + w/2 and y + h/2 (that will be the center of the detected object). If there was other information that you needed, feel free to call me on Discord: moises_dias.
Caraaca irmão, como eu n tinha pensando nisso antes! Nunca mais tinha trabalhado no meu projeto pq o delay n deixava ser util. Agora com esse algoritmo de predict vou tentar aplicar e ver se funciona! Vlwwzão
Yoo, great work you doing here, need a help how can I make different actions to every simple detections, I mean if In my game its detecting "Ork" and "Troll" I want it to make different actions when it detects "Troll" and different when detects "Ork"
hi there! thanks for the feedback! so, if you want to take different actions based on the class of the detection you should do the following: when you call the function "improc.proccess_image(ss)" you will receive a list with all the detections, every item in the list is a dictionary with 5 keys: x, y, w, h and class_name. you can see in my video that i am filtering only the class fruit in this part of the code: coordinates = improc.proccess_image(ss) coordinates = [c for c in coordinates if c["class_name"] == "fruit"] so basically, use this atribute (the "class_name" key in the items of the list) in a if else statement to perform diferent actions based on the class of the detection. i hope it helps but if you have any other question feel free to ask!!
nice video bro. I followed the video and got the weights, but now I am so confused about how to use those weights in the game. Should I make a python code? if yes, how do I get it to work with those weights?
Hello! you can use the same logic as presented on this video to do that, but changing from mouse movement and click to keyboard key press. You can see more functionalities of Pynput on the documentation: pypi.org/project/pynput/
Eu consegui detectar perfeitamente os mobs. Porém, eu não faço ideia de como eu coloco para rodar. Eu colei o código que fiz do ragnarok e dei run, mas não acontece nada. Se alguém puder dar uma força, dá um toque. Prometo não passar em branco
opa, bom dia! então, muito provavelmente o que está ocorrendo é que o anti-cheat do ragnarok (gepard shield) está detectando que o movimento do mouse e o click são simulados por um programa, e ai esse anti cheat não considera essa ação.. passei pelo mesmo problema :(
@@moisesdepaulodias7980 Ahh entendi. Tipo, quando eu rodo meu código python no VSC, ele funciona, eu utilizo a biblioteca opencv. Só que não detecta muito bem. Pelo jeito, vou ter que usar esse mesmo XD. Muito obrigado viu.
Hi There! Now I am working on another series about machine learning with Google Cloud, but maybe in the future I will do that, thanks for the sugestion!!
Hello, thank you for this video on how to automate bot! I have a question tho, are all your code coded using Python? If yes, I am gonna make myself learn it for this purpose 😂. I currently know some basics of Javascript, and I am going to learn Python anyways in the future, again, thank you!
Hi there, I will think about it ;) I'm thinking of creating an object detector without using machine learning, using methods such as color detection, shape recognition, and template matching. I might use Subway Surfers as a basis to try out these methods.
muito bom meu parceiro, agora duvido você fazer um novo video lançando pra como colocar isso no arduino e como pegar as box de personagens inimigos em jogos de tiro tipo SOT (Sea of thieves) quero destruir todos aqueles desse jogo, me mandaram foto do pinto do nada com uma mensagem de bom dia, reclamei no discord deles e eles disseram "ele só vende conteudo, tem problema não" fiquei imaginando se fosse uma criança, AGORA ME AJUDE A DESTRUIR ELES
Hi there! Hufus, you can use the same logic that I use in the video and change mouse movement to key press. In this case, I had to move the mouse to the fruits and then left-click and drag. But you can, for example, move the mouse to an enemy and then press a keyboard key, using keyboard.press instead of mouse.press.
@@moisesdepaulodias7980 but for plataform games like maplestory! this works and everything i try using chatGPT to change and make the bot go a bit close to start attacking but the was soo many errors no mather how i try to fix or change or modify and the server im on have like a protection and only works with autohotkey and run with UI access!! and i just want to know how to make a script that moves my char and go up and down and press E if see a mob or make the bot go in circles like using teleport "right+w" will teleport left all the keys+w will teleport and thats it! i manage to to kida do that in AHK and a bit on python but its a mess it doest press up+w 1 times it presses way to fast even if i put sleep, 1000 i mean it doest go in circles some times it goes more than 3 times and break the circle path i wnat to go! i dont know if i explain my self right!
We are still waiting for the Vertex AI video!! Your content is awesome, please keep it coming!!
Hey, thank you so much for your comment! It really means a lot to me!!! I started a new job a few months ago and I don't have too much free time at the moment, but I will make the Vertex AI tutorial as soon as I can!!
@@moisesdepaulodias7980 I was wondering if you had a chance to get the next video done?
Thank you!
I would love to see more from you.
Hi there, right now I am kinda busy but I really intent to do more content like this, thanks for the kind words!!
I'm subscribed, and eagerly waiting for your future videos!
Your videos are amazing! It's really helping alot with understanding what is all going on. Thank you!
Hello ,author, your work is excellent, I would like to ask, when you do the labeling work in Makesenseai after taking screenshots to obtain data, do you need to label the fruit that has been cut?
Can't wait for the next video 😍
hey thanks man, I am kinda busy at the moment but I want to make new videos soon, thanks for your comment!
Muito bom o tutorial Moises. Curti também o video anterior! To na missão de fazer um "jogador automatico" pra buscar item pra mim no Ragnarok, e o seus videos estão me ajudando muito!
Thanks for the tutorial, very interesting to watch
Pretty good, thank you for sharing with us.
Fiz o modelo no tutorial do seu video passado mas usei poucas imagens e ficou inconsistente. tem Como eu adicionar mais imagens ao modelo atual ou deveria fazer outro do zero?
Thanks for the tutorial. I begin to learn Your video is very good. Thank you for making a good video. ❤
Thank you for your comment!!
Hey, I like the subject, I havera great video suggestion for you, make a new video about How to connect a Yolov8 Custome trained model to Connect toAny Game Window. In your first video you have been using Yolov4 , the current version is V8 which has so many options, for Detection ,Tracking, Segmentation and so on, now it is even easier deploying implementing for custom projects.
Hi there, thanks for your suggestion! I used v4 because I had used it in the past, but when I have the time, I intend to do an updated version with v8!
Make tutorial for character that use autopath, example in diablo 2. How can i make my character follow a specific long path. and walk in that specific long path only from location-A to location-B. Thank you! Pls make a guide 🙏
im saved to my list now , thanks
Hi!
What if an object is detected, move the mouse to the center of the first detected object. which code should I delete/change? Thanks
Hello! It will be hard to explain this in a youtube comment, but basically, in the variable "new_coordinates" you have a list with all the detected objects, each object being an dictionary with keys x, y, h, w and class_name. basically, you need to use the x, y, h and w to get the coordinate and move the mouse to the first object, the center of the first detected object will be:
(new_coordinates[0][x] + new_coordinates[0][w]/2, new_coordinates[0][y] + new_coordinates[0][h]/2)
hi, can you use this method to farm item at ragnarok online?
bem melhor que muitos tutorial brasileiros.
much better than many Brazilian tutorials
cara, muito obrigado mesmo pelo elogio!!!! fico muito feliz que você tenha gostado!!
Hi sorry i'm kinda new to this, so i don't really understand everything. I would like to do the same with a 3D combat game, but by making the agent attack and dodge. Can i do this with that technology? or can i mix YOLO and Pynput in order to do that ? Thanks a lot
Hi there! Well, you can do that using YOLO and Pynput, but it will be very challenging. In my case, I only had to detect fruits, but in your case, you will need to not only identify your opponent, but when your opponent is going to attack you so you can dodge.
For your purpose, if you have the opportunity to try a packet sniffer (a program that intercepts the communication between a client and a server) to detect your opponents' actions, instead of using an image detector, I highly recommend starting there.
hi I still use Bluestacks to test different games, but the mouse pointer doesn't point to Bluestacks itself, but to the detected window. What could be the possible reason?
Hello, I recommend you move the Bluestacks window to the top left corner of your screen. The mouse movement considers this position as the starting point. Could you try it and let me know if it worked?
Can it be extended to auto aim? I really want to try to Re7.
Sure! basically, in my video i moved the mouse when i detected a fruit, and in your case you must move the mouse (and maybe press the left mouse button to shoot) when you detect a specific monster in re7
@@moisesdepaulodias7980 thank you 👍👍
make more of these videos please
Thanks, man!! I will make more videos when I have the time. I believe that my next videos will be about how you can use object detection without machine learning because for tasks like detecting objects with simple 2D graphics (such as pixel art), there are a lot of other ways to do that using OpenCV.
hi, the window "DETECTED OBJECTS" renders a blackscreen, i have tried to launch bluestacks in admin but it doesn t change anything. Do you know what can i do?
Hi there. I had the opportunity to try to help one person with the same problem in a Discord call, and I believe that this is related to some anti-cheat of your game. Because in this Discord call, the code worked in some games, but in others it generated black screens (I also suspect that games within Steam may generate black screens as well).
I suggest you try with other games to see if that is your case. For me, it worked with the games Terraria, Ragnarok Online, Diablo 2, and Fruit Ninja using the BlueStacks emulator.
I am trully sorry I can't help any further, but I didn't encounter this problem while creating the tutorial.
Brother great explanation in simplest way loved you videos, but can you tell me what code to write if i want the crosshair in game to go to the custom trained object"s center. Hope to hear from you soon, waiting for your future videos.❤❤❤❤
Hello! I don't know if I understand correctly, but if you want to get the center of the detected object (x, y), you can just get the return of the function improc.process_image(ss) (that will return a list with the detected objects), and for each one, you can get the x + w/2 and y + h/2 (that will be the center of the detected object). If there was other information that you needed, feel free to call me on Discord: moises_dias.
Caraaca irmão, como eu n tinha pensando nisso antes! Nunca mais tinha trabalhado no meu projeto pq o delay n deixava ser util. Agora com esse algoritmo de predict vou tentar aplicar e ver se funciona! Vlwwzão
opa, imagina cara!! fico feliz que o video foi útil para você Matheus!!!
good video. thanks bro
Yoo, great work you doing here, need a help how can I make different actions to every simple detections, I mean if In my game its detecting "Ork" and "Troll" I want it to make different actions when it detects "Troll" and different when detects "Ork"
hi there! thanks for the feedback!
so, if you want to take different actions based on the class of the detection you should do the following:
when you call the function "improc.proccess_image(ss)" you will receive a list with all the detections, every item in the list is a dictionary with 5 keys: x, y, w, h and class_name. you can see in my video that i am filtering only the class fruit in this part of the code:
coordinates = improc.proccess_image(ss)
coordinates = [c for c in coordinates if c["class_name"] == "fruit"]
so basically, use this atribute (the "class_name" key in the items of the list) in a if else statement to perform diferent actions based on the class of the detection.
i hope it helps but if you have any other question feel free to ask!!
Thanks a lot that's what I was looking for !
And that class of detection = label which I giving in labeling stage on previous tutorial Right?
?
hello mate, how can I contact you?
Insano mano, parabéns!
Good job man , amazing
nice video bro.
I followed the video and got the weights, but now I am so confused about how to use those weights in the game. Should I make a python code? if yes, how do I get it to work with those weights?
Can I make the character move? W forward S back A go slow D go right Let the characters in the game walk to find the items themselves.
Hello! you can use the same logic as presented on this video to do that, but changing from mouse movement and click to keyboard key press. You can see more functionalities of Pynput on the documentation: pypi.org/project/pynput/
Eu consegui detectar perfeitamente os mobs. Porém, eu não faço ideia de como eu coloco para rodar. Eu colei o código que fiz do ragnarok e dei run, mas não acontece nada. Se alguém puder dar uma força, dá um toque. Prometo não passar em branco
opa, bom dia! então, muito provavelmente o que está ocorrendo é que o anti-cheat do ragnarok (gepard shield) está detectando que o movimento do mouse e o click são simulados por um programa, e ai esse anti cheat não considera essa ação.. passei pelo mesmo problema :(
@@moisesdepaulodias7980 Ahh entendi. Tipo, quando eu rodo meu código python no VSC, ele funciona, eu utilizo a biblioteca opencv. Só que não detecta muito bem. Pelo jeito, vou ter que usar esse mesmo XD. Muito obrigado viu.
Hi can u make a video how to navigate a character on a map?
Hi There! Now I am working on another series about machine learning with Google Cloud, but maybe in the future I will do that, thanks for the sugestion!!
Keep going ❤
hey there! Thank you for the encouragement!! I will try to keep making videos about this kind of automation.
Hello, thank you for this video on how to automate bot! I have a question tho, are all your code coded using Python? If yes, I am gonna make myself learn it for this purpose 😂. I currently know some basics of Javascript, and I am going to learn Python anyways in the future, again, thank you!
Hi there! thanks for your comment!
And yes, I only used Python on this tutorial (and a little bit of bash on the training step).
Cool man!! can u make tutorial for noobs like me on subway surfers
Hi there, I will think about it ;)
I'm thinking of creating an object detector without using machine learning, using methods such as color detection, shape recognition, and template matching. I might use Subway Surfers as a basis to try out these methods.
Great video!!!
Que isso cara, um vídeo melhor que o outro. WTF
Muito bom o conteúdo, parabéns.
muito bom meu parceiro, agora duvido você fazer um novo video lançando pra como colocar isso no arduino e como pegar as box de personagens inimigos em jogos de tiro tipo SOT (Sea of thieves) quero destruir todos aqueles desse jogo, me mandaram foto do pinto do nada com uma mensagem de bom dia, reclamei no discord deles e eles disseram "ele só vende conteudo, tem problema não" fiquei imaginando se fosse uma criança, AGORA ME AJUDE A DESTRUIR ELES
hi can you please do a video but to farm and kill mobs? :) thankyou boss
can u make pls with the keys insted of mouse???
Hi there! Hufus, you can use the same logic that I use in the video and change mouse movement to key press. In this case, I had to move the mouse to the fruits and then left-click and drag. But you can, for example, move the mouse to an enemy and then press a keyboard key, using keyboard.press instead of mouse.press.
@@moisesdepaulodias7980 but for plataform games like maplestory! this works and everything i try using chatGPT to change and make the bot go a bit close to start attacking but the was soo many errors no mather how i try to fix or change or modify and the server im on have like a protection and only works with autohotkey and run with UI access!! and i just want to know how to make a script that moves my char and go up and down and press E if see a mob or make the bot go in circles like using teleport "right+w" will teleport left all the keys+w will teleport and thats it! i manage to to kida do that in AHK and a bit on python but its a mess it doest press up+w 1 times it presses way to fast even if i put sleep, 1000 i mean it doest go in circles some times it goes more than 3 times and break the circle path i wnat to go! i dont know if i explain my self right!
Thanks a Lot