AI-For-Road-Safety

Challenge hosted by Omdena Hyderabad Chapter

Original Repo Link : https://github.com/OmdenaAI/omdena-india-roadsafety

Final Presentation Link : https://docs.google.com/presentation/d/1Gk53mIN270ovEqfSlX6FUAC4_ynEWU3hHXVGxUJZtRg/edit?usp=sharing

Google Drive : https://drive.google.com/drive/folders/1cGs5Inm8AaTxmG9Ad6-lX6X_5CWBq7Zx

Data Wrangling & Pre-Processing Dashboard : https://public.tableau.com/app/profile/nikhil2069/viz/road_safety_EDA/EDAofTweets

Computer Vision Demo Link : https://drive.google.com/file/d/153NX6Sm9Re-gc8ZLYQ7GuJrxWhM7YRhp/view?usp=sharing

Streamlit Application :

Streamlit Link : https://share.streamlit.io/prathimacode-hub/ai-for-road-safety/main/main.py

Docker Application :

Docker Link : https://hub.docker.com/repository/docker/snikhil17/omdena_road_safety

Docker Steps:

Note:

• snikhil17 : login ID of Dockerhub (use own when deploying)
• omdena_road_safety: name of the image created (you may use something else)

To Push the image and run in your PC

Create the image

• docker build -t snikhil17/omdena_road_safety .

Run the image in local PC

• docker run -it -p 8501:8501 snikhil17/omdena_road_safety
• To run streamlit: once docker image is run. Open a new browser and run http://localhost:8501/

To Push image in docker-hub

• docker login
• docker push snikhil17/omdena_road_safety

To pull the image and run in your PC

• docker pull snikhil17/customer_intention_1:latest
• docker run -it -p 8501:8501 snikhil17/omdena_road_safety
• To run streamlit: once docker image is run. Open a new browser and run http://localhost:8501/

Tasks Involved:

Natural Language Processing

Label & Sentiment Generator - Check out this application through Streamlit Demo App Link provided.

Computer Vision

Eye Gaze Estimation + Drowsiness Detection + Yawn Detection

This application doesn't support Streamlit, hence it can compiled directly on Local Drive using generated Driver Attention Estimation .exe file files uploaded in the drive.

Demo Link : https://drive.google.com/file/d/153NX6Sm9Re-gc8ZLYQ7GuJrxWhM7YRhp/view?usp=sharing