Summry

Overwhelmed by video lectures before exams? Use Summry!

Comment

💡 INSPIRATION 💡

Because of the COVID pandemic, many lectures and meetings are now being recorded and delivered to students as video lectures. These lectures can be long and not interactive, making them difficult to understand and retain information. Summry was created to address this issue by turning lectures into concise, searchable notes.

⚙️ WHAT IT DOES ⚙️

Summry creates one-page study guides from hour-long videos, allowing students to review material more easily, thoroughly, and efficiently.

🛠️ HOW WE BUILT IT🛠️

We used a variety of tools and techniques to build Summry, including scene detection with Numpy & OpenCV, OCR, and speech recognition with Google Cloud Speech-To-Text API.

😣 CHALLENGES WE RAN INTO 😣

Completing such a large project in just 30 hours was a challenge, as was learning about natural language processing. We also encountered issues with using the Google Cloud API and understanding how different forms of punctuation affected the results of speech transcription.

🎉 ACCOMPLISHMENTS WE ARE PROUD OF 🎉

We are proud to have created a solution to a problem we personally faced, as well as being able to utilize and learn about multiple APIs and Python modules. For many of us, this project also provided our first exposure to more advanced machine learning techniques.

What we learned

Through this project, we learned about teamwork and task delegation, as well as web development frameworks and how to create a user-friendly interface. We also gained experience with Google Cloud's APIs, flask, multi-processing and multi-threading.

What's next for Summry

In the future, we plan to include timestamps with summary snippets for easier reference, as well as links to external study resources. We also hope to enable text editing after speech transcription.

Share this project:

Updates