Gym-routine-extractor

Pitch for Gym Routine Extractor

Inspiration: The increasing popularity of home workouts and online fitness videos inspired us to create Gym Routine Extractor. We noticed that while there is a wealth of information on YouTube, it can be challenging for users to follow along or remember complex workout routines. We wanted to make extracting and following workout routines from videos more accessible and efficient.

What it does: Gym Routine Extractor takes a YouTube workout video as input, transcribes the content, and extracts a structured workout routine, complete with exercises, sets, reps, and special notes. It provides an easy-to-understand, step-by-step routine that users can save, revisit, or use for tracking their workouts.

How we built it: We used OpenAI's Whisper for video transcription and fine-tuned language models to extract relevant workout details from the transcript. Our application is built with Python, leveraging LangChain to streamline natural language processing and summarization. We also incorporated a user-friendly interface to make interacting with the tool intuitive and efficient.

Challenges we ran into: One of the biggest challenges was accurately extracting specific exercise information from varying video formats. Workout videos often use informal language, making it difficult for traditional text processing to identify key exercises and details. Additionally, working with YouTube's API for video data extraction and ensuring consistency in the results required significant fine-tuning.

Accomplishments that we're proud of: We're proud to have developed a solution that makes following online workouts easier and more organized. Successfully combining transcription, natural language processing, and structured output into an intuitive product is something we’re very proud of. Users can now extract meaningful, actionable workout plans from their favorite fitness videos, all in a matter of seconds.

What we learned: We learned a lot about working with large language models, how to effectively process natural language from informal content, and how to integrate various APIs to create a cohesive tool. We also learned the importance of understanding user needs and iterating based on feedback to improve the experience.

What's next for Gym Routine Extractor: We plan to expand the tool to support multiple languages, making it accessible to a broader audience. We also want to integrate personalization features, allowing users to adjust workout intensity based on their fitness level. Lastly, adding workout tracking and progress analysis would make this a comprehensive personal training companion.

Built With

• fast-api
• gpt
• langchain
• openai-whisper
• python