Yolov5 Tutorial - What is YOLOv5 Object Detector (Real-Time Object Detection)

what is yellow v5 yellow v5 is a popular real-time object detector it is a pytorch implementation of the yellow single shot detector known for its blazingly fast speed and very good accuracy the name yolo stands for you only look once as the name suggests yolo is a single stage detector that processes the input image once to detect objects it is fundamentally different from two-stage object detectors like faster rcnn the yellow family of detectors from yellow v1 to v4 is implemented in a deep learning framework called darknet it is written in c yellow v5 on the other hand builds on top of a pie torch implementation of yolo v3 as you may know pytorch is a popular super cool deep learning framework backed by facebook that is one of the reasons why yolo v5 is so popular let's look at some other reasons yellow v5 contains a collection of five object detection models the models are all summarized in this table the first column shows the number of model parameters or the model size the second column shows the accuracy and the third and fourth columns show the time it takes to process a single image on a cpu and on a gpu the smaller models are less accurate but very fast as the model size increases so does the accuracy but they are also slower let's go over these models one by one yellow v5n where m stands for nano is the smallest in the family and it is meant for edge mobile and iot solutions it is also supported by the opencv dnn module and it is tiny it is 2.5 megabytes in int 8 format and about four megabytes in fp32 which is floating point 32 format eulo v5s where s stands for small has about 7.2 million parameters and is ideal for running inference on the cpu yellow v5m where m stands for medium has 21.2 million parameters it provides a good balance between speed and accuracy and therefore it is well suited for many data sets and applications yellow v5l where l stands for large has 46.5 million parameters it is ideal for data sets where we need to detect small objects euro v5x where x stands for extra large is the largest among the five models with 86.7 million parameters it has the highest accuracy but it also is the slowest eulo v5 also provides easy ways to export the models into different formats you can export it to both pi torch and tensorflow like formats and if you love javascript you would be pleased to know that it can also be exported to tensorflow.js format people using intel cpus gpus vpus or even the opencv ai kit can export the model to openvino format if you use nvidia hardware like jetson you can easily export it to tensorrt format and finally the highly portable onyx onnx format is also supported yolo v5 makes exporting all these formats very very easy now let's look at logging training an ai model can take hours sometimes days so we need to check if the training is going well by logging the current model's performance there are two popular tools for logging and visually checking the model's performance the first one is tensorboard and the second one is weights and biases both of them are supported by yellow v5 both are great tools but many people love weights and biases but to use weights and biases you must create an account and provide the api credentials before starting the training for proper logging you need none of that for tensorflow all the logs are automatically saved by yellow v5 so in summary yellow v5 is an excellent object detector that gives you a collection of pre-trained models suitable for varying speed and accuracy demands the logging features are great and you can export the models to all popular formats that's all i wanted to cover in this video this is sacha malik signing off your guide to the fascinating world of computer vision and ai thank you

Channel: LearnOpenCV

Views: 15,775

Keywords: Artificial Intelligence, AI Basics, AI for Beginners, Machine Learning for Beginners, Data Science, Machine Learning for Python, Deep Learning, YOLO, You Only Look Once, YOLOv5, Object Detection, Object Detector, Real Time Object detection, YOLO Explained, Yolo Tutorial, What is YOLO, What is YOLOv5, YOLOv5 Explained, YOLO SSD, PyTorch, Single Stage Object Detection, YOLOv5s, YOLOv5n, YOLOv5m, YOLOv5x, YOLOv5l, Custom Object Detection Training, logging, Visualization, yolov5 tutorial

Id: JzHNIcvpGk8

Length: 5min 16sec (316 seconds)

Published: Mon Apr 25 2022