PlantSnap

Inspiration

Climate change has the ability to affect every aspect of our lives, from the availability of food and water to the stability of economies and societies. As young leaders we wanted our project to help the environment and bring a simple solution to a problem that is not commonly addressed.

What it does

Our project PlantSnap takes an image of a plant and identifies whether its one of the 5 most common invasive plant species, and informs the user. Through technology, we can make this task that is normally done by educated individuals, accessible to everyone.

How we built it

We found a model from HuggingFace which was an image classifier. We changed the model to identify 5 of the top most common invasive plant species in North America. Then we used PyCharm to take a picture as a .jpg file and input it to run it through the model and return a result to the user. The for loop checks the probability of the inputted image matching to one of the 5 invasive species. If the probability score is less that 50% for all 5 plants, then it returns a result saying that there isn't a match. Otherwise, it returns a result with the highest probable invasive plant from the 5 and the probability score.

Challenges we ran into

As complete beginners to Machine Learning we ran into many challenges. Firstly, coming up with an idea that would implement machine learning. We brainstormed many ideas and ultimately chose PlantSnap. When looking for databases in the beginning we were having trouble finding a database with invasive species. Eventually we found a model on HuggingFace that was an image classifier, which was very helpful. Our main challenge was figuring out how a user could input an image to try PlantSnap and we eventually decided to use PyCharm to do this.

Accomplishments that we're proud of

We were very proud of completing our second hackathon, learning about machine learning for the first time, and completing a project on top of that. We had no previous knowledge on what machine learning was even about, and how useful it is. It was very rewarding to learn so much in a short amount of time.

What we learned

We learned what Machine Learning is for the first time and its many applications. We attended the workshops. We learned about HuggingFace and how we can easily search for databases and models that are already trained. We will definitely use these skills in the future.