Inspiration
RepWise was born from a simple yet powerful need — to make proper exercise form accessible to everyone, anywhere. Many people face challenges when it comes to maintaining correct posture during workouts. Professional trainers can be expensive and hard to access, while poor form often leads to preventable injuries. Many users also feel uncertain about whether they are performing exercises correctly, and most fitness apps fail to offer real-time, personalized feedback.
We wanted to change that. Our vision was to create a system that blends computer vision and*AI-driven analysis*= into a digital personal trainer that is always available, affordable, and capable of offering instant, actionable feedback.
What It Does
RepWise provides real-time feedback on your movements, helping users perfect their form, reduce injury risk, and make every rep count.
Its key features include:
Real-Time Form Analysis – RepWise uses Yolov8 pose detection to track body movements and analyze over 23 different exercises, including push-ups, squats, and deadlifts. It provides instant visual and audio feedback, letting users know immediately whether their form is correct.
Accomplishments We’re Proud Of
We’re proud of achieving seamless integration between computer vision, AI, and voice processing without sacrificing real-time performance. RepWise maintains smooth, lag-free video feedback while the AI assistant operates in parallel.
We also designed a user-friendly interface that provides meaningful visual feedback without overwhelming users, making the experience both informative and intuitive.
What We Learned
Building RepWise taught us how to balance complex real-time processes while maintaining usability and stability. We gained valuable experience in threading, resource management, and designing AI systems that feel natural and responsive to human interaction.
What’s Next for RepWise
Our next steps include expanding RepWise’s AI-driven capabilities to provide personalized workout recommendations, injury prevention insights, and adaptive difficulty adjustments based on each user’s performance. We also plan to enhance the voice assistant’s emotional awareness and add more detailed analytics to help users track long-term progress.
Log in or sign up for Devpost to join the conversation.