YOLOv4 Object Detection - WebApp using Streamlit📸

Object Detection using coco.names dataset , weights and configuration files of real time object detection algorithm YOLOv4

Demo :

This is WebApp of my previous repo here

To change the weights and configurations file , you may do so by changing the file directory of the same.

Demo on Streamlit

Demo.mp4

Requirements 🏫

- aiortc==1.2.1
- av==8.0.3
- matplotlib==3.4.2
- numpy==1.21.1
- opencv-python-headless==4.5.3.56
- pydub==0.25.1
- streamlit>=0.84.2
- streamlit_webrtc>=0.25.0
- typing_extensions==3.10.0.0
- openpyxl

COCO Dataset 🍫

S.No	Sports	Living	Things	Vehicles	Safety	Food	Dining	Electronics
1	frisbee	bird	bench	bicycle	traffic light	banana	bottle	tvmonitor
2	skis	cat	backpack	car	fire hydrant	apple	wine glass	laptop
3	snowboard	dog	umbrella	motorbike	stop sign	sandwich	cup	mouse
4	sports ball	horse	handbag	aeroplane	parking meter	orange	fork	remote
5	kite	sheep	tie	bus	Nil	broccoli	knife	keyboard
6	baseball bat	cow	suitcase	train	Nil	carrot	spoon	cell phone
7	baseball glove	elephant	chair	truck	Nil	hot dog	bowl	Nil
8	skateboard	bear	sofa	boat	Nil	pizza	microwave	Nil
9	surfboard	zebra	pottedplant	Nil	Nil	donut	oven	Nil
10	tennis racket	giraffe	bed	Nil	Nil	cake	toaster	Nil
11	Nil	person	diningtable	Nil	Nil	Nil	sink	Nil
12	Nil	Nil	toilet	Nil	Nil	Nil	refrigerator	Nil
13	Nil	Nil	book	Nil	Nil	Nil	Nil	Nil
14	Nil	Nil	clock	Nil	Nil	Nil	Nil	Nil
15	Nil	Nil	vase	Nil	Nil	Nil	Nil	Nil
16	Nil	Nil	scissors	Nil	Nil	Nil	Nil	Nil
17	Nil	Nil	teddy bear	Nil	Nil	Nil	Nil	Nil
18	Nil	Nil	hair drier	Nil	Nil	Nil	Nil	Nil
19	Nil	Nil	toothbrush	Nil	Nil	Nil	Nil	Nil

Usage 👥

# Paramaters which can be tuned to your requirements
confThreshold = 0.5
nmsThreshold = 0.2

# for reading all the datasets from the coco.names file into the array
with open("coco.names", 'rt') as f:
    class_names = f.read().rstrip('\n').split('\n')
    
# configration and weights file location
model_config_file = "yolo-config\\yolov3-tiny.cfg"
model_weight = "yolo-weights\\yolov3-tiny.weights"

How to Run this Program using streamlit🏃‍♂️

streamlit run main.py

Reference 🧾

You can read more about YOLO and more about streamlit

About Me 😉

I am Rahul Arepaka, II year CompSci student at Ecole School of Engineering, Mahindra University

Feel free to edit with the source code and enjoy coding

Contact 📞

You may reach me using

• Mail 📧
• Linkedin 😇