Smartwatches, fitness trackers, smart home gym equipment, and smartphone applications are just a few of the technological innovations engineers and computer scientists have developed to improve fitness training experiences in recent years. Deep learning algorithms, in particular, can significantly advance these tools to more effectively cater to specific users’ demands.
Researchers from the University of Brescia in Italy recently developed a computer vision system for a smart mirror. This technology could increase fitness training efficiency in home and gym settings. This method is built on a deep learning algorithm that has been trained to identify human gestures in video recordings, according to a report published by the International Society of Biomechanics in Sports. The video can be found here.
A skeletonization technique is used in the low-cost computer vision system created by Lanza and his coworkers, which is running on an integrated Nvidia Jetson Nano device with two fisheye cameras. The two fisheye cameras were used to record video for the study, and the researchers trained the system to analyze and recognize human motions in the video.
One of the system’s advantages is that there are no objects in contact with the user.
The ideal smart mirror would be able to assess fitness workouts in a manner equivalent to or even superior to human personal trainers. It might, for instance, track the number of repetitions users complete for particular exercises while also identifying the basic motion (such as traction, flection, rotation, etc.) of various body parts. It displays all fitness-related data that the mirror detects and calculates, updating in real-time so that users may monitor it while working out or use it to boost their training efficiency. In a series of tests, Lanza and his colleagues assessed the performance of their computer vision system with an emphasis on its capacity to monitor and forecast users’ fitness levels as they performed biceps curls. The researchers discovered that their low-cost vision system could provide useful fitness-related data when users did straightforward workouts for fitness with well-designed and calibrated software. The innovative solution could considerably assist users’ training without a supervising coach in home and gym locations when linked with AB-smart Horizon’s mirror.
By examining unprocessed body kinematic data, the intelligent assessor should be able to interpret qualitative information. Therefore, before training this model, the researchers will first gather much data through fitness tests with athletes and less-experienced fitness trainees.
This Article is written as a research summary article by Marktechpost Staff based on the research paper 'DEEP LEARNING FOR GESTURE RECOGNITION IN GYM TRAINING PERFORMED BY A VISION-BASED AUGMENTED REALITY SMART MIRROR'. All Credit For This Research Goes To Researchers on This Project. Check out the paper and reference article. Please Don't Forget To Join Our ML Subreddit
Prathvik is ML/AI Research content intern at MarktechPost, he is a 3rd year undergraduate at IIT Kharagpur. He has a keen interest in Machine learning and data science.He is enthusiastic in learning about the applications of Machine learning in different fields of study.