Alphashion

A picture of our home page
A picture of our camera page
A picture of our graph page
Our bar graph(clothing type vs. quantity)
Our line graph(price of clothes over time bought)
Our pie chart
Our weather tab that recommends clothing
Selecting a picture of shirt to analyze
A picture taken by our camera tab
A picture of our Kivy main file
A picture of our machine learning using TensorFlow

Inspiration

Summer. Days of scorching heat, and unpredictable weather. Sunny skies quickly darken to thunderstorms, which bring cooler weather with them. Global warming only makes this weather even more fickle. Summer also brings new clothes, fashion, and bags to the table. Many people, including us, with overflowing closets and too many clothes to keep track of, struggle with keeping track of what to wear and bring around, especially in the face of unpredictable weather. From shorts to cargo pants, hiking backpacks to purses, and flannels to tank tops, there's so much to keep track of. The people need an easy to use app that allows them to store information about their clothes and bags. We built Alphashion to solve this problem.

What it does

Alphashion allows the user to either take pictures from the camera tab, or manually select an image from the device. The user then can upload both the image and other specific information about the clothes onto the "cloud" to keep their "closet" clean and organized. During this uploading process, our app will analyze the image and decide which type of clothing it is using machine learning. Furthermore, based on the data the user has entered, Alphashion is able to make clothing recommendations for the user based, on local temperature which is obtained by a GPS function. Alphasion also provides statistical analysis and visual graphs for the users to get a sense of how much they spend on buying different types of clothes, how many of each type of clothes they have currently, and the percentage of each type of clothes one has in one's virtual closet.

How we built it

We used python and TensorFlow for the backend and processing. We manually trained a machine learning model to determine what type of clothing is in the pictures taken. We also used matplotlib to display graphs. For the frontend, we used Kivy and a little bit of tkinter.

Challenges we ran into

Our ideas were very creative and useful but too ambitious. We did not have enough time to implement all of them. However, we plan to implement some of these later. Training the model to be more accurate was also very difficult. Another thing that challenged us was stitching the front end and back end together.

Accomplishments that we're proud of

We are proud of being able to stitch all the front end and back end together. We have a working app that is able to generate weather predictions and beautiful graphs. We are also proud about improving the accuracy of our AI, and creating a select image function. Furthermore, this was one of our teammates first time using python so we are proud of succeeding nonetheless.

What we learned

We learned that good projects require time, skills, and good planning. We also learned a lot of skills such as Kivy, Tensorflow machine learning, and matplotlib. We also learned how hard stitching together parts to a whole can be, as we all divided up the work and had to piece it together at the end. We learned how to work and communicate together as a team.

What's next for Alphashion

We plan to fully develop Alphashion and publish it on the app store. We also plan to pitch this idea to a company. We also want to further implement more functions and ideas, and fix the minor bugs we currently have. We want to polish the app up and further improve the accuracy of the AI so that we can use this to help our everyday lives. We hope that one day, Alphashion will help make people's lives easier.