Idea Name: |
Traffic-sign Recognition |
Slogan: |
Your Safety, Our Priority. |
Supervisor Name: |
Shahzor Ahmad |
Supervisor Designation: |
Assistant Professor |
Supervisor School: |
CEME |
Supervisor Department: |
Electrical Engineering |
Contact number: |
+92 336 1572440 |
Email ID: |
[email protected] |
Abstract: |
Our Traffic Sign Recognition project aims at automatic detection and classification of traffic signs in local environment using Deep Convolutional Neural Networks. |
What is the unmet need in society that your idea will fulfill ? |
Reverting driver's attention back to road alerting him with the latest road sign warning so that he could timely act accordingly, is very crucial to road safety and is an issue that still needs attention. Pakistani manufactured automobiles need to incorporate this feature for enhanced road safety. |
Who needs it ? How many would benefit ? |
Can subsequently be used in Advanced Driver's Assistance System (ADAS), or for help in navigation of self driving cars at later stages. Every individual driving a locally manufactured vehicle can benefit from it. This could be additionally appended to individual vehicles as well. |
How will the solution works |
Our Deep CNN will be trained on GTSRB data set of 50,000 images, its accuracy will then be tested and the algorithm will be subjected to testing in local environments. Hardware implementation on NVIDEA Jetson Nano be done in the final stages for real time applications. Performance will be enhanced by error correction using transfer learning. |
Who are your competitors ? How is your solution different |
No such competitors in the local market yet, which makes it a unique idea/product at this level. |
Status: |
new |
Entry Date & Time: |
2021-01-14 (1521) |