一款能计算重复次数并评估Apple Watch肌肉力竭的科技
2 分•作者: baraa_bilal•大约 1 个月前
大家好，我开发了这个产品。
Riven 可以将 Apple Watch 的运动数据转化为三项信息：一组训练的开始和结束时间、你正在进行的锻炼项目、完成的次数，以及目标肌肉接近力竭的程度。它仅使用 Apple Watch 内置的 100 Hz 运动传感器，无需任何手动输入。告别手动记录。
其背后的理念由来已久且经过充分验证：当接近力竭时，动作的完成速度会减慢。基于速度的训练（Velocity-based training）多年来一直沿用这一原理，但通常需要一个安装在杠铃上的传感器或一个对准杠铃的摄像头，因此主要局限于杠铃训练。我利用 Apple Watch 内置的传感器实现了相同的概念。
最具挑战性的部分是判断一组训练的开始和结束，尤其是在人来人往的健身房中活动时，以及区分有效次数和干扰信号。经过大量的机器学习训练，我们已经将其优化到非常出色的水平。
App Store：https://apps.apple.com/app/riven-reps-muscle-failure/id6770362701
网站：https://riven.fit
非常乐意回答任何问题或听取您的反馈！
查看原文
Hi HN, I built this.<p>Riven turns Apple Watch motion into three things: when a set starts and ends, which exercise you're doing, how many reps you did, and how close that muscle got to failure. The only sensor it uses is the Watch's 100 Hz motion sensor. No manual input at all. Say goodbye to manual logging.<p>The idea behind it is old and well-tested: reps slow down as you approach failure. Velocity-based training has used this principle for years, but it typically requires a sensor on the bar or a camera pointed at it, so it's mostly limited to barbell exercises. I built the same concept using only the sensors already inside the Apple Watch.<p>The tough part is knowing when a set starts and finishes, especially when you're moving around a busy gym, and differentiating reps from noise. After extensive ML training, we've gotten it to a point where it works really well.<p>App Store: https://apps.apple.com/app/riven-reps-muscle-failure/id6770362701
Website: https://riven.fit<p>Happy to answer any questions or hear your feedback!