Inspiration
- The lack of waste management practices has caused significant issues.
- One major concern is overwhelming landfills, leading to environmental degradation and health hazards.
- Valuable resources are wasted due to lack of awareness about recyclable materials.
- Improper disposal methods often lead to valuable resources being wasted, causing significant environmental consequences.
- Improper disposal practices contribute to air and water pollution, posing severe health risks to nearby communities.
- The absence of recycling and waste-to-energy initiatives hampers efforts to combat climate change and rely on alternative energy sources.
- The failure to segregate, recycle, and dispose of waste properly has triggered serious issues like overflowing landfills, resource depletion, pollution, and climate change.
- Urgent action is required to implement efficient waste management practices and pave the way for a more sustainable future.
What it does
- TrashTrove provides a user-friendly platform that utilizes advanced image recognition technology.
- Users can determine the recyclability of a product by uploading an image to the platform.
- The platform swiftly analyzes uploaded product images to identify their recyclability status.
- For recyclable products, comprehensive information about recycling methods is provided to encourage environmentally friendly choices.
- Non-recyclable products are guided towards appropriate disposal methods to minimize environmental and health harm.
- Responsible waste management is essential for mitigating pollution, reducing landfill burden, and combatting climate change.
How we built it
- ML Model was trained using TensorFlow, Keras, Fastai, Torch, NumPy
- Frontend was built with React and Vite with Bootstrap
- Backend was made using Flask
- AI search was done using GPT-3 making use of OpenAI API
Challenges we ran into
- One of the main difficulties we faced was ensuring the Algorithm Accuracy. This means how well the system can recognize and correctly tell if a product is recyclable or not from the uploaded images.
- We faced was the lack of free computing resources. We had limited server capacity, which made it impossible to host the website without compromising its performance.
Accomplishments that we're proud of
- Our proudest accomplishment in this project is developing a highly accurate and user-friendly waste management platform.
- By leveraging image recognition technology, we enable users to make informed decisions about waste disposal with an accuracy of 95%.
- We are also proud of fostering awareness and promoting responsible waste management practices.
- Dataset consisting of that 22000+ images was used for training the ML Model.
- We successfully hosted locally the backend in spite of the lack of free resources.
What we learned
- Importance of Waste Management: We have gained a deeper understanding of the critical importance of effective waste management practices.
- Impact of Recycling Awareness: Lack of awareness about recyclable materials has significant repercussions on resource conservation and sustainability.
- Environmental Consciousness: Through this project, we have gained a heightened sense of environmental consciousness.
- User-Centric Design: Developing a user-friendly interface has taught us the significance of user-centric design. A simple and intuitive interface encourages greater user engagement and participation in responsible waste management practices.
What's next for Trashtrove
- Our next step for TrashTrove is to develop a mobile app that takes waste management convenience to the next level. With the app, users will no longer need to upload images; instead, they can simply scan the product using their smartphone's camera.
- By focusing on user convenience, accuracy, and global collaboration, we are dedicated to creating a more sustainable and environmentally conscious future for all.
- We plan on making our waste segregation more specific in future for better and optimized results.
- To motivate users to use the application, we will be bringing some kind of return-reward system based on the frequency of their usage.


Log in or sign up for Devpost to join the conversation.