We propose a near real-time solution for frame-rate enhancement that enables the use of existing sophisticated pose estimation solutions at elevated frame rates. Our approach couples a keypoint human pose estimator with optical flow using a multistage system of queues operating in a multi-threaded environment. As additional contributions, we propose a pose tracking solution and an approach to overcome errors caused by optical flow. A reduction in error in the range of 30–35% is observed at practical frame rates of pose estimator (4–6 frames per second) while processing $$1920\times 1080$$ 1920 × 1080 resolution 30 frames-per-second videos at native frame rate. Slower frame rates have increased the reduction of error up to 50%, thereby promoting the use of cheaper hardware and sharing of expensive hardware. Thus, while improving accuracy by enabling sophisticated pose estimation models to operate at above-par frame rates, our approach reduces cost per frame by promoting efficient resource utilization.

中文翻译：

---

实时二维人体姿态估计的跟踪和帧率增强

我们提出了一种近乎实时的帧速率增强解决方案，该解决方案能够以提高的帧速率使用现有的复杂姿态估计解决方案。我们的方法使用在多线程环境中运行的多级队列系统将关键点人体姿态估计器与光流相结合。作为额外的贡献，我们提出了一种姿态跟踪解决方案和一种克服光流引起的误差的方法。在位姿估计器的实际帧速率（每秒 4-6 帧）下观察到误差减少 30-35%，同时处理 1920 美元\times 1080 美元 1920 × 1080 分辨率、每秒 30 帧的视频原生帧率。较慢的帧速率将错误减少率提高了 50%，从而促进了更便宜硬件的使用和昂贵硬件的共享。因此，