Alireza Delisnav的动态

?? Thrilled to Share My Recent OpenCV Assignment! ?? I've been diving into an exciting project utilizing OpenCV for powerful computer vision solutions. Here's what I've been up to: Image Processing Mastery: Implementing techniques like filtering, edge detection, and color space transformations to enhance and analyze images. Dynamic Object Detection: Developing systems to detect and track objects in real-time, enhancing applications like surveillance and autonomous driving. Facial Recognition: Creating an accurate face recognition system using OpenCV and integrating it with machine learning models for improved identification. Seamless Integration: Bringing these capabilities together into a cohesive application that demonstrates the practical uses of computer vision. Performance Tuning: Optimizing algorithms and processes to ensure efficient and high-speed operation across various scenarios. This project has significantly boosted my expertise in computer vision and deepened my understanding of OpenCV’s powerful toolkit. I'm excited to take these skills forward and continue pushing the boundaries of what's possible with computer vision technology! #AIMERS #AIMERSOCIETY #APSCHE #DATASCIENCE #OPENCV #HUGGINGFACE #ARTIFICIALINTELLIGENCE #MACHINELEARNING #DEEPLEARNING #BIGDATA #COMPUTERVISION #NATURALLANGUAGEPROCESSING #AI #TECHNOLOGY #INNOVATION #ROBOTICS #AUTOMATION

Hi?? Thrilled to share my experience working on a groundbreaking project: Brain-Controlled 3D Avatar Drone Navigation System! Over the past year, I developed an innovative system that enables real-time control of a 3D avatar drone using brain wave inputs, combining advanced technology with human cognition. ???? ?? Key Highlights: BCI Headset Integration: Captured and processed brain wave data from a BCI (Brain-Computer Interface) headset, which served as the primary input dataset for controlling drone movements. Neural Network Implementation: Developed neural networks to interpret these brain wave patterns accurately, ensuring precise control over drone functions such as takeoff, landing, and directional movements. Python & Machine Learning: Utilized Python to implement machine learning algorithms, enabling accurate signal processing and real-time responses. Docker Deployment: Employed Docker for seamless deployment, ensuring the system's scalability and efficiency in various environments. GUI Development: Designed and implemented a user-friendly interface (GUI) to facilitate easy interaction and control, incorporating core drone functionalities. ?? A special thank you to Professor Dr. #AdrianoCavalcanti for their valuable mentorship and allowing me to work on this project as a graduate assistant. #BCI #AI #Python #TensorFlow #MachineLearning #DroneTechnology #NeuralNetworks #Innovation #SCSU

?? Vision Eye with Distance Calculation for Vehicles ?? Using?Ultralytics?YOLOV8 Featuring: ???Object mapping and distance estimation prevent collisions by gauging proximity to surrounding objects. ???Object mapping and distance estimation prevent collisions by gauging proximity to surrounding objects. ???Optimizing traffic flow, these technologies adjust signals and routes based on real-time data, enhancing safety and efficiency. ?? For Collaboration and Consultancy projects, feel free to DM us We would love to hear your thoughts! Stay in the loop with the updates by following Antar AI. OpenCV OpenCV University Muhammad Rizwan Munawar Piotr Skalski Roc4Tech Arnaud Bastide Alessandro Ferrari Ayush Chaurasia #computervision??#objectdetection?#objectTracking?#yolo?#yolov8?#opencv?#roboflow #DistanceCalculation?#VisionEye?#ai?#Ultralytics

Project Exhibition 2 ?? Ever wondered how traffic monitoring works behind the scenes? Let me take you through a sneak peek into our latest project! ?? Project Overview: We've been working on a cutting-edge traffic monitoring system employing computer vision techniques to detect vehicles in real-time from video feeds. This is a project which we built during our college project exhibition. ??? Tech Stack: Our system utilizes OpenCV, a powerful library for computer vision tasks. We've implemented background subtraction and contour detection algorithms to identify vehicles accurately. ?? Video Processing: We preprocess each frame by converting it to grayscale and applying Gaussian blur to enhance vehicle detection accuracy. ?? Vehicle Detection Logic: Our vehicle detection module employs a sophisticated approach to detect vehicles crossing a specified line. We track vehicle counts, distinguishing between 2-wheelers and 4-wheelers, contributing to traffic analytics. ?? Objective: Our goal is to provide insightful data for traffic management, enabling better decision-making for urban planning and infrastructure development. ?? Key Features: Real-time vehicle detection and counting. Classification of vehicles based on size (2-wheelers vs. 4-wheelers). Dynamic tracking of incoming and outgoing vehicles. #TrafficMonitoring #ComputerVision #SmartCities #Innovation #AI #OpenCV

Dear LinkedIn community, I want to share some of the recent projects I've been involved in during my tenure as a postdoctoral fellow at the Mobile Robotics Lab (MRL), McGill University. These projects merge advanced research with real-world applications, covering state estimation, perception, control, and planning. ?????????????????????? ???????????? ???????????????? ?????????????? ??????????-???????? ?????????????????????????? ???????????????? Our research, a segment of which was presented at CRV2023, focuses on leveraging model-free reinforcement learning for underwater object tracking. Departing from traditional PID controllers, we implemented a centralized RL agent to control the robot, ensuring robust performance. We utilized classic controllers for safe exploration of the state-action space. Our approach involved extensive simulation studies in Unity and real-world deployment on Nvidia Jetson boards. Check out the paper and GitHub repository: Paper1: https://lnkd.in/eQyqHPdr, Paper2: https://lnkd.in/e9W78S9F, Code: https://lnkd.in/e-6GrRUf ???????????????? ???????????????????? ?????????? ?????????? ???????????????? ???????????? ???????? ???????????? ???????????????????????? Presented at ISER2023, our research investigates the integration of large language models (LLMs) into navigation planning. We leveraged Lseg, a large vision-language model, to bypass the need for extensive dataset collection and model training. By incorporating insights from GPT-3.5, we achieved enhanced success rates by constraining terrain types, landmarks, and adverbs. Simulation studies conducted in Unreal Engine validated our approach. Explore more in the paper and GitHub repository: Paper: https://lnkd.in/eCj47cpQ, Code: https://lnkd.in/eC_gDNx7 ????????????????????????-?????????? ???????????? ???????????????? ???? ?????????????????? ?????? ?????? ??????????-?????????? ???? ?????? ?????????????? ???????????????????? Scheduled for presentation at ICRA2024, this research delves into an innovative hybrid paradigm for offroad navigation. By modeling the environment and vehicle dynamics, we optimized traversal through diverse terrain types while prioritizing collision avoidance and managing trade-offs between speed and uncertainty. Paper: https://lnkd.in/e6n47qGq, Code: https://lnkd.in/eG8XDzGg Our experimental studies encompassed field trials at Bellairs Research Center, Barbados, and Toronto University Camp near Gull Lake. I want to express my deepest gratitude to Professor Gregory Dudek for his invaluable guidance and support throughout these projects. I'm also immensely thankful to the dedicated members of the MRL lab, including Khalil Virji, Farnoosh Faraji, Andrew Holliday, Lucas Berry, Travis Manderson, Nikhil Kakodkar, Jean-Fran?ois Tremblay, and others who I might have missed. Your contributions have been instrumental in the success of these endeavors. Here is a quick video showcasing our project results. Stay tuned for more updates!

?? Rescaling Images and Videos with OpenCV????? In the field of computer vision, image and video rescaling is a fundamental operation for optimizing performance, reducing memory usage, or simply standardizing input sizes for machine learning models. With OpenCV, rescaling is both efficient and straightforward. ?? Rescaling Images with OpenCV: Rescaling an image involves adjusting its width and height while preserving the overall structure. This is often necessary when preparing datasets or visualizing data on different devices. ?? Rescaling Videos with OpenCV: For video processing, resizing frames is crucial when working with large video files or real-time applications. OpenCV makes it easy to rescale each frame during video capture, helping to reduce the computational load. ? ?? Key Applications: - Optimizing model performance: Rescale input images and video frames to standard dimensions to improve efficiency in AI and machine learning tasks. - Reducing memory and bandwidth: Ideal for applications in streaming, where processing large video frames may be costly. - Preprocessing in real-time systems: Autonomous systems, drones, and robotics often need rescaled video for faster decision-making. Scaling images and videos efficiently is essential when building robust, scalable computer vision systems. With OpenCV, handling such tasks becomes intuitive, making it an indispensable tool for anyone working in AI, robotics, or image processing. #OpenCV #ComputerVision #Rescaling #AI #MachineLearning #ImageProcessing #VideoProcessing #Python #AutonomousSystems #Robotics

?? New Project Using AI and Parking Management! ?? I'm really excited to share a recent project I've developed using OpenCV and video processing. ?? ?? Project Goal: The project aims to analyze parking lots and determine whether there are available spaces for cars by automatically and intelligently analyzing video footage. ?? ?? How Does It Work? With the use of surveillance cameras, the system monitors parking lots and analyzes visual signals to accurately identify available and occupied spaces. ? A detailed report is continuously generated in a TXT file, providing real-time data on the number of available and occupied parking spaces. ?? ?? Why Is This Project Important? This project represents a significant step towards automating parking management and increasing efficiency in finding empty spaces, reducing the time spent searching for a parking spot. ?? I’m looking forward to applying these smart solutions in more areas to enhance our daily experiences through AI! ?? ?? If you're interested in more details or want to discuss how to apply this project, feel free to reach out! #AI #OpenCV #VideoProcessing #MachineLearning #ComputerVision #ParkingManagement #Automation #SmartCities #Innovation ??

Hello Connections ?? ?? Excited to share my latest task on OBJECT TRACKING ! ?? Using a pre-trained dataset and I downloaded the best.pt model for object detection. I’ve implemented real-time object tracking with a webcam???.I used YOLOv8 for detecting objects like birds,smily yellow balls and diseased leaf. ?? Integrated the best.pt model into my PyCharm project, enabling real-time object tracking. ?? Using the webcam, It tracks various objects, capturing and labeling their precise x, y coordinates. It was amazing experience and enhanced my skills in model training and inferencing?. #ObjectTracking #MachineLearning #ComputerVision #AI #DeepLearning #PyCharm #AIMERS #AIMERSOCIETY #APSCHE #DATASCIENCE #OPENCV #HUGGINGFACE #ARTIFICIALINTELLIGENCE #BIGDATA #NATURALLANGUAGEPROCESSING #TECHNOLOGY #INNOVATION #ROBOTICS #AUTOMATION

Photogrammetric modeling of the city. We apply the deep learning car recognition algorithm during the 3D modeling procedure, thus creating a city model without moving cars. This work greatly reduces the time cost of tedious manual editing work. www.get3d.ai, #programmetry #citymodeling #DeepLearning

?? Exciting Project Update: Road Pothole Detection Using YOLOv8 Custom Segmentation! ?? I’m thrilled to share my latest computer vision project, which focuses on real-time road pothole detection using a YOLOv8 custom segmentation model. This innovative solution has the potential to greatly enhance road safety by identifying potholes in video footage with precision. ?? Key Highlights: Powered by YOLOv8 for detecting and segmenting potholes with high accuracy. Uses OpenCV and cvzone to visualize pothole contours and label the affected areas in real-time. Custom-trained model fine-tuned to detect road damage efficiently. Real-time visualization of potholes, providing essential data for road maintenance teams to take action. This project is a step toward making our roads safer and contributing to the prevention of vehicle damage and accidents due to road hazards. ?? How it works: The model processes video frames, detecting potholes and highlighting them with bounding boxes and contours. Detected areas are labeled with real-time feedback, providing clear insights into road damage conditions. I'm eager to explore further improvements and real-world applications for this solution, especially in cities that struggle with road maintenance issues. ?? Curious to see it in action? Feel free to check out the code and the project on my GitHub: [https://lnkd.in/dUYnp8dd]. Looking forward to your thoughts and suggestions! Let’s discuss how AI can revolutionize road safety! ???? #AI #ComputerVision #YOLOv8 #DeepLearning #RoadSafety #PotholeDetection #OpenCV #MachineLearning #Innovation #RealTimeDetection #Transportation #RoadPotholeDetection #YOLOSegmentation #AIForTransportation