Skip to content

The basic object detection model using Pretrained Model and Transfer Learning.

Notifications You must be signed in to change notification settings

abhishekpwr/real-time-pothole-detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

38 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Real Time Pothole Detection

This is the project of real time object detection model, using pretrained weights of volov5 algorithm, that performs real time Pothole Detection.

Click on YouTube icon below to see the real time video inference or Click Here.



Dataset Used

In this Project I have used custom labelled dataset with some images from google to make the learning process more complex..

Training Dataset Sample

Training Results

Yolo vs Efficientdet performance

YOLOv5 Nano models

Pretrained Checkpoints for YOLOv5 transfer learning

Model size
(pixels)
mAPval
0.5:0.95
mAPval
0.5
Speed
CPU b1
(ms)
Speed
V100 b1
(ms)
Speed
V100 b32
(ms)
params
(M)
FLOPs
@640 (B)
YOLOv5n 640 28.4 46.0 45 6.3 0.6 1.9 4.5
YOLOv5s 640 37.2 56.0 98 6.4 0.9 7.2 16.5
YOLOv5m 640 45.2 63.9 224 8.2 1.7 21.2 49.0
YOLOv5l 640 48.8 67.2 430 10.1 2.7 46.5 109.1
YOLOv5x 640 50.7 68.9 766 12.1 4.8 86.7 205.7
YOLOv5n6 1280 34.0 50.7 153 8.1 2.1 3.2 4.6
YOLOv5s6 1280 44.5 63.0 385 8.2 3.6 16.8 12.6
YOLOv5m6 1280 51.0 69.0 887 11.1 6.8 35.7 50.0
YOLOv5l6 1280 53.6 71.6 1784 15.8 10.5 76.8 111.4
YOLOv5x6
+ TTA
1280
1536
54.7
55.4
72.4
72.3
3136
-
26.2
-
19.4
-
140.7
-
209.8
-

In this project I have used YOLOv5l for transfer learning.

Suggest Improvements

If you need the dataset or the trained weights of this model, you can contact me through email or Linkedin! This project is in its early stage and I belive it will get improved over the time in future iterations. Please let me know if you have any suggestions. Thank you!

Thank You :)


About

The basic object detection model using Pretrained Model and Transfer Learning.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages