- Видео 11
- Просмотров 5 704
Alzaib Karovalia
Добавлен 14 янв 2012
Robotic Arm to Pick and Drop a Load
Course Project
MSE 312: Mechatronics Design II
Simon Fraser University
MSE 312: Mechatronics Design II
Simon Fraser University
Просмотров: 36
Видео
Self-Driving Car Using Artificial Intelligence
Просмотров 3963 года назад
Simon Fraser University MSE Capstone Project Team Arcane AI
GTA-Self driving vehicle (without object detection algorithm)
Просмотров 8634 года назад
This is the first version of the autonomous vehicle with joystick steering control, without any object detection algorithm.
Autonomous Vehicle in GTA-5 Using Deep Learning and YOLOv3
Просмотров 2,9 тыс.4 года назад
The following is the updated version of the project I have been working on for a while. The goal is to build a self-driving car in GTA-5 with deep learning and computer vision, which can navigate in different environments. This self-driving car is trained with a custom dataset of 39,046 images, which is artificially expanded using image augmentation, to predict the steering angle and throttle. ...
Self-driving AI bot in GTA-5 (ResNet-50 VS NVIDIA's PilotNet)
Просмотров 4994 года назад
I've been working on creating a self-driving AI in GTA-5 for the last few weeks. This project is inspired by Sentdex's implementation. However, I used GoogleColab and Keras to train each network and experimented with different parameters to produce optimal results. Both of the following networks, ResNet-50 and NVIDIA's PilotNet, are trained using the dataset uploaded by Sentdex on his website. ...
YOLOv3-tiny Real Time Detection
Просмотров 854 года назад
Real Time Object Detection using YOLOv3 on opencv deep neutral network module. Only center part of the video is being processed to increase the speed of real time detection.
MSE 310: BYTRONICS
Просмотров 875 лет назад
This system separates plastic and metal and assembles them. It also kicks out the unassembled objects. Done in partnership with Junayed Ahmed Nowshad.
Hydroelectric Generator (Prototype) Course Project
Просмотров 1136 лет назад
Hydroelectric Generator (Prototype) Course Project "Applied Science, Technology and Society"
Simon Fraser University, MSE 110, Line Maze Solving Robot
Просмотров 2537 лет назад
Line Maze solving robot, that returns to the original position by calculating the shortest path (avoiding the loop). This robot is made using one sensor only so ignore the adjusting part.
bro im also getting this error can u help me with this error plz Traceback (most recent call last): File "C:\Users\F\Desktop\Autonomous-Self-Driving-Car-GTA-5-master\test_model_steer.py", line 162, in <module> main() File "C:\Users\F\Desktop\Autonomous-Self-Driving-Car-GTA-5-master\test_model_steer.py", line 89, in main outs = net.forward(output_layers) NameError: name 'output_layers' is not defined >>>
It seems like I might have deleted "output_layers = [layer_names[i[0] - 1] for i in net.getUnconnectedOutLayers()]" when cleaning the code. Please see the following link to implement yolo detection. github.com/Alzaib/Traffic-Signs-Detection-Tensorflow-YOLOv3-YOLOv4/blob/main/detection_from_folder.py
Nice
Nice car
Automatic car (Btw) very nice
Nice car
when im running the model im getting this error Traceback (most recent call last): File "C:\Users\F\Desktop\Autonomous-Self-Driving-Car-GTA-5-master\test_model_steer.py", line 162, in <module> main() File "C:\Users\F\Desktop\Autonomous-Self-Driving-Car-GTA-5-master\test_model_steer.py", line 42, in main model = load_model(model_name) File "C:\Users\F\AppData\Roaming\Python\Python39\site-packages\keras\saving\save.py", line 206, in load_model return saved_model_load.load(filepath, compile, options) File "C:\Users\F\AppData\Roaming\Python\Python39\site-packages\keras\saving\saved_model\load.py", line 109, in load meta_graph_def = loader_impl.parse_saved_model(path).meta_graphs[0] File "C:\Users\F\AppData\Roaming\Python\Python39\site-packages\tensorflow\python\saved_model\loader_impl.py", line 113, in parse_saved_model raise IOError( OSError: SavedModel file does not exist at: steer_augmentation.h5\{saved_model.pbtxt|saved_model.pb} any fixes
Download the training weights from here: drive.google.com/drive/folders/1laagsAkn_TqjyKw1zvd5Okct5Sd3WONF and save them in same folder as test_model_steer.py Also make sure you have all the libraries (keras, tensorflow OpenCV etc) installled
bro what is the resolution i should run the game and where should i place it
bro if i put a waypoint on map will i t drive to it
I didn't trained it to follow the waypoints, but if you recollect the data by driving to the waypoints, and retrain the network, I think it will reach the waypoint. Something like this: arxiv.org/pdf/1811.10119.pdf
thank u : )
@@mynamejeff5193 Hey did you re-train the script to make it go to waypoints? If so can you please share it to me because I do not understand the coding.
Foooooooòd
I love it
Bhot acha
Hello, very interesting project. I would like to know in brief how did you extract the keyboard presses and how did you align them with the video in order to make the training set
Look at the tutorial by Sentdex. I used a similar approach in this project. Instead of using keypresses from the keyboard, I recorded steering and throttle information from Xbox joystick. Check out this detailed article I wrote for this project: medium.com/@alzaibnasiruddin/building-a-self-driving-vehicle-in-gta-v-using-deep-learning-and-convolutional-neural-network-696b38b4c81e
I have no idea about this stuff. But I know Ure Proportional gain is very high! But chah Gaye ho bro.
This is the iteration without object detection, so for each frame, the AI predicts steering angle and throttle which is passed into a virtual joystick (vJoy) that drives the vehicle. Do check the final version integrated with object detection (YOLOV3) which performs significantly better than this one: ruclips.net/video/pKTgLk3KeiA/видео.html
@@playlist422 But in reality the self driving car knows the steering angle. Is it something u must do because it is GTA or just something ure exploring. FYI I've seen all ure videos. This one was newer (date posted) so I figured another update! What about Gagan? I know he got into ML,DL and all.
@@omarceentayyab this is something I'm still exploring. I've read some articles where a self driving cars are integrated with much complex networks to produce best result. I've also seen few projects where steering is controlled with PID controller.
Would you mind sharing it? Do you have discord?
Here's an excellent tutorial by Sentdex: ruclips.net/channel/UCfzlCWGWYyIQ0aLC5w48gBQ pythonprogramming.net/game-frames-open-cv-python-plays-gta-v/ I am still working on this project by adding longitudinal control with some object detection algorithm. I will share everything once I'm done.
great work
Good video ! I made a line following robot too!