Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196309,
        author = {Shreya Tale and Sanket Wahane and Sahil Mandiya},
        title = {Artificial Intelligence(AI) Based Interactive Fitness Trainer},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {3066-3073},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196309},
        abstract = {Due to the global spread of sedentary lifestyles, which has been made worse by the digitalization of work and the post-pandemic shift toward remote living. Although digital fitness platforms have made instructional content more accessible to a wider audience, they have historically been unable to offer real-time, corrective feedback on biomechanical form. This shortcoming raises the risk of injury and decreases the effectiveness of training [9]. The architectural layout, theoretical underpinnings, and implementation plan for an interactive fitness trainer powered by artificial intelligence (AI) are presented in this research paper. The suggested system uses consumer-grade hardware without specialized graphical processing units (GPUs Accessible, scalable, and individualized physical fitness interventions are desperately needed) to perform high-fidelity, monocular human pose estimation (HPE) using the cutting-edge MediaPipe BlazePose framework. The system incorporates geometric heuristics for precise repetition counting, a finite state machine (FSM) Many people are exercising without professional supervision as a result of the growing popularity of at-home fitness regimens, which frequently leads to poor posture and a higher risk of injury. The majority of fitness apps currently in use rely on pre-recorded videos and don't provide real-time interaction or helpful feedback. This paper presents an AI-based Interactive Fitness Trainer that analyzes user posture in real time using camera input and provides corrective feedback using computer vision and machine learning techniques. The proposed system uses MediaPipe BlazePose for pose estimation, Flask or Streamlit for an interactive user interface, and machine learning algorithms such as Decision Tree, Logistic Regression, and Random Forest for exercise recognition. Platform-neutrality, affordability, and accessibility are the goals of the system [5].},
        keywords = {Artificial Intelligence, Fitness Trainer, Pose Estimation, MediaPipe, Computer Vision, Machine Learning, Real-Time Feedback},
        month = {April},
        }