Implementation of an Intelligent Real-Time Squat Detection and Performance Evaluation System Using Computer Vision and Pose Estimation Techniques

References

[1] Z. Cao et al., “OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 43, no. 1, pp. 172–186, 2020.

[2] G. Chen et al., “AI-Based Posture Correction for Fitness Applications,” Sensors, vol. 21, no. 7, pp. 2564–2572, 2021.

[3] V. Bazarevsky et al., “BlazePose: On-device Real-time Body Pose Tracking,” Google Research Blog, 2021.

[4] G. Bradski, “The OpenCV Library,” Dr. Dobb’s Journal of Software Tools, 2000.

[5] Google Research, “MediaPipe: Cross-Platform Framework for Machine Learning Pipelines,” Google AI Blog, 2022.

[6] P. Maji et al., “Human Action Recognition Using Vision-Based Techniques,” Journal of Visual Communication and Image Representation, 2022.

[7] G. Van Rossum, “Python Programming Language,” Python Software Foundation, 2023.

[8] F. Camarena et al., “An Overview of the Vision-Based Human Action Recognition Field.” Mathematical and Computational Applications, 28(2), 61, 2023.

[9] F. Shafizadegan et al., “Multimodal Vision-Based Human Action Recognition Using Deep Learning.” Multimedia Tools and Applications, 2024.

[10] P. S. JosephNg, “A Cost-Effective Approach Using CNN and LSTM Models.” SSRN, 2025.

[11] D. A. Phalke, “Artificial Intelligence in Fitness: Pose Estimation and Movement Correction.” Cureus, 2025.

[12] V. Bazarevsky et al., “BlazePose: On-device Real-time Body Pose Tracking.” arXiv, 2020.

[13] I. Grishchenko et al., “BlazePose: On-device Real-time Body Pose Tracking.” Google Research Blog, 2021.

[14] W. Supanich et al., “Machine Learning-Based Exercise Posture Recognition System Using MediaPipe Pose Estimation Framework.” ResearchGate, 2023.

[15] S. Pal et al., “Review-Based AI-Driven Posture Correction and Personalized Fitness Assistant Using Computer Vision and Augmented Reality.” ResearchGate, 2025.

[16] A. Tharatipyakul, “Deep learning-based human body pose estimation in …” ScienceDirect, 2024.

[17] K. E. Appiah et al., “A Mobile-Phone Pose Estimation for Gym-Exercise Form Anomaly Detection.” VISAPP 2024, 2024.

[18] D. A. Phalke et al., “Artificial Intelligence in Fitness: Pose Estimation and Movement Correction.” Cureus, 2025.

[19] Ali K. AlShami et al., “SMART-Vision: Survey of Modern Action Recognition Techniques in Vision.” arXiv, 2025.

[20] 20.Jungpil Shin, Najmul Hassan, Abu Saleh Miah, Satoshi Nishimura, “A Comprehensive Methodological Survey of Human Activity Recognition Across Diverse Data Modalities.” arXiv, 2024.

[21] Nature article efficient human activity recognition on edge devices using TinyML. 2025.

[22] Application of human activity/action recognition: a review. Multimedia Tools and Applications, 2024.

[23] Machine Learning for Human Activity Recognition: State-of-the-Art. MDPI, 2024.

[24] A systematic literature review on human activity recognition. Springer, 2025.

[25] HARNet in deep learning approach—a systematic survey. Scientific Reports, 2024.

[26] A Survey on AI-Powered Exercise Posture Correction System. IJAEM, 2024.

[27] Ai-Powered Gym Posture Optimization. IRJMETS, 2024.

[28] AI-Based Posture Correction, Real-Time Exercise Tracking, and Feedback using Pose Estimation Technique. IEEE CCIS, 2024.

[29] Human Pose-based Estimation, Tracking and Action Recognition with Deep Learning: A Survey. arXiv, 2023.

[30] A Survey of Video Action Recognition Based on Deep Learning. Knowledge-Based Systems, June 2025.

@article{1712786,
      author = {S. Ranjithkumar, Dr. G. Nallavan, S. Sugumar},
      title = {Implementation of an Intelligent Real-Time Squat Detection and Performance Evaluation System Using Computer Vision and Pose Estimation Techniques},
      journal = {Iconic Research And Engineering Journals},
      year = {2025},
      volume = {9},
      number = {6},
      pages = {1107-1112},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1712786.pdf},
      abstract = {The growing integration of artificial intelligence (AI) and computer vision in fitness and healthcare has enabled new possibilities for real-time movement tracking and automated performance evaluation. This research introduces AI Squat Sense, an intelligent, real-time squat detection and assessment system designed to monitor and enhance exercise performance. The system employs OpenCV for video acquisition and image processing, and MediaPipe Pose for extracting skeletal landmarks. By analyzing joint angles at the hip, knee, and ankle, AI Squat Sense accurately determines squat depth, counts repetitions, and evaluates posture quality. A graphical user interface (GUI) developed using Tkinter and ttkbootstrap ensures ease of use, offering real-time visual and auditory feedback through optional modules like pyttsx3 and simpleaudio. Experimental results demonstrate over 95% detection accuracy under normal lighting with minimal computational delay. The proposed system provides a cost-effective, accessible, and robust solution for fitness tracking, rehabilitation, and sports performance optimization, bridging the gap between personal fitness tools and intelligent motion analysis technologies.},
      keywords = {Artificial Intelligence, Computer Vision, Pose Estimation, Exercise Monitoring, Fitness Tracking, Human?Computer Interaction, Real-Time Analysis, OpenCV, MediaPipe.},
      month = {December},
      doi = {https://doi.org/10.64388/IREV9I6-1712786}
  }