YOLOv8 | Fire Detection on a Custom Dataset using YOLOv8

[Music] viewers welcome to the pirisa channel so in this video we will going to the fire detection with you love yet so let's start one create simple report after we are a dead data set compiling in PT format you can see the where it is already given so ultralysis will be format of the YOLO V8 so uh here the result uh we are going to the YOLO directory you can see it so we add detection model sorry directory my English is not good I am trying to my best so how to run this algorithm and my target is uh create the audience to achieve so best so my Internet is slow okay no problem we are going to the entire internet so you can see the uh one mp4 file predict file and the testing file uh first of all you download it from here also going to the git command use this command copy and paste go to the folder uh I'm giving the dependency requirements so uh go to the directory and run this command I'm already having got my desktop version so you can see the desktop fire Direction you look weird so weird induction you can see the test file uh yeah it's status file first of all I will show you how to run and you can see the python dependency python 3.9 Point writing so see the desktop CD YOLO here the fire reduction man file okay so we are going to the directory of ultralysis YOLO we get reduction part so all script will be here uh near the Britain wet fire reduction here the file of the you can see the now I'm mentioning the video One video simple video this woman will show you the two fire candles and here's the one picture you can see it so first of all you go to the here predict that thank you so you can see 8.0.0 version of the alternative and video will be trend so this video will be Trend and run here the prediction part so this result will be sure after video we are going to the Target image source image name is will show you can see the fire.jpg okay fire.jpg and that Trend number four okay run detection train number four you can see the result of the file prediction uh 76 percent is accurate okay so we are going to the now this file will be tested uh so python test dot file you see the real time result no detection part yeah 72 percent will be this deck so it's taking time so my Mac version it's not gpus it's CPU best we are testing in the CPUs so this woman will be fired will be detect you can see it anymore so it's depend on you a lot of people's first of all train the video you I will mention it how to Trend and uh then also I mentioned you how to Rel Temple deck in video format and it's already short and locks you also see the logs okay uh you want to see the second day I want to see that second demo of the video so download one simple file detection video you can download from here you download from here then you download from here go to put your word directory and we will expand the simple format of the euroviet here the library's dependence libraries here the one class detector form after drag the pre-processing for format which is attributes of the dividing Rule and frame per second will be divided 16 by 32 then post the frame how to post a frame means you're just like a pre-process and write the result what you want to write of your frame just like your labeling imagine here this video will be showing you how to detect the far fire and accuracy 0.76 means 70 percent accuracy so here the class of the model we get yellow V8 so checking the size of the image and close them last function okay so we are going to the now test format uh this is scribble also will be test if the libraries computer vision Library both your model return model I will put in my git account so Link in description so you get the results and predict the video You also mentioned that your Source camera just like a zero one and show in the logs printer logs okay you see the logs okay our video we will download we can go and you can see the rename is renew test dot MP4 thank you so Source name will be change test dot MP4 why we cannot show the real time I will mention here python test dot pi you get the result so you see the 72 percent accuracy fire detected and 70 80 percent now will be accurate so I hope you like it please support our channel for more getting information updates any type of confusion comment down

Channel: Pyresearch

Views: 4,815

Keywords: Education, Pyresearch, Pyresearch - YouTube, how to Grow, money, facebook app, facebook, how to set goal, yolo object detection algorithm, yolo ai, yolo algorithm explained, yolo algorithm github, yolo object detection tensorflow, yolo darknet, yolo machine learning, yolo tutorial, yolo detection, yolo deep learning, yolo image recognition, object detection, darknet yolo, mask rcnn, mask r-cnn, faster rcnn, mask rcnn deep learning, rcnn, video mask rcnn, mask rcnn pytorch

Id: roQ8J8ggP44

Length: 9min 59sec (599 seconds)

Published: Thu Feb 02 2023