Inspiration
Small business owners are often in the market looking for vacant storefronts. The business owners might like how a spot looks or get excited about a busy street corner, but they may not consider other important factors that go into this decision making process. As commuter students, we took Tech Square as our inspiration to analyze storefront visibilities. Over the weekend, we worked on a tool that essentially removes a huge chunk of the guesswork out of this whole process.
What it does
VisiFy is storefront evaluation tool powered by AI that assigns a combined, comprehensive score based on a combined score of visibility analysis and suitability of businesses. This would help our business clients to understand the strength of their location, evaluating visual obstructions and assessing multiple viewpoints. We analyze the storefronts to give prospective business owners and entrepreneurs solid, data-backed reasons to choose one location over another.
How we built it
We leveraged the following datasets and APIs-
Tom Tom API Google Maps Street View API US Cenus Demographic Market Data
Tech stack used - Python, React, HTML, JavaScript, A-Frame.
We also utilized a YOLOv8 Object Detection Model to train some and leverage detect visibility of storefronts.
Challenges we ran into
Some of our members came in without any background in data science or analytics, but we leveraged nice datasets and models to get accurate object detection models. The utilization of computer vision also seemed tricky, as it required large datasets that needed, with some being unstructured and some needing to be trained.
Accomplishments that we're proud of
With minimal background in data science and analytics, we leveraged datasets and models to get accurate object detection models. Another accomplishment we are proud of is the leveraging of Google Street View API and the seamless integration between the flask and the React front end.
What we learned
As mentioned before, all of our members came in with little to no experience with data analysis or data science in general. We learned a lot this weekend. From leveraging Google Maps Street view API to implementing an AI chatbot, this event was full of learning and celebration of knowledge.
What's next for VisiFy
The next steps for VisiFy include training a more computationally intensive model for our clients. One of the steps we have already identified is the integration of additional APIs and datasets. Precisely speaking, we are interested in using more data sources, such as satellite data for enhanced demographic, geographic and traffic insights . We also want to improve our web interface and machine learning enhancement, to analyze business trends while improving accuracy. Due to these reasons and more, we believe our project has great scalability.
Log in or sign up for Devpost to join the conversation.