Put Your Muscle Into It: Introducing XEM 2 , a Novel Approach for Monitoring Exertion in Stationary Physical Exercises Leveraging Muscle Work Conference: ; ; 6166 CCS: Human-centered computing CCS: Human-centered computing Human computer interaction (HCI) CCS: Human-centered computing Interactive sy

Jana Franceska Funke, Mario Sagawa, Georgious Nurcan-Georgiou, Naomi Sagawa, Dennis Dietz, Evgeny Stemasov, Enrico Rukzio, Teresa Hirzle · 2026

Read Put Your Muscle Into It: Introducing XEM 2 , a Novel Approach for Monitoring Exertion in Stationary Physical Exercises Leveraging Muscle Work Conference: ; ; 6166 CCS: Human-centered computing CCS: Human-centered computing Human computer interaction (HCI) CCS: Human-centered computing Interactive sy by Jana Franceska Funke, Mario Sagawa, Georgious Nurcan-Georgiou, Naomi Sagawa, Dennis Dietz, Evgeny Stemasov, Enrico Rukzio, Teresa Hirzle (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-15472.txt.

Introduction

Jana Franceska Funke and seven co-authors present a new way to track effort in exercise in a 2026 arXiv paper. Their system is called XEM 2, short for exercise exertion muscle-work monitor. It uses a camera to measure muscle work during stationary exercise. The method rests on the Hill-type muscle model, a classic model of how muscles produce force.

The team aims to fix weak spots in older ways of tracking effort. Calorie counts can be imprecise when movements are not uniform. Scales based on self-perception say little about strain in each body part. XEM 2 instead shows muscle work directly, part by part. That gives users and trainers a clearer view of where the effort lands.

Catalog record

CREATORJana Franceska Funke, Mario Sagawa, Georgious Nurcan-Georgiou, Naomi Sagawa, Dennis Dietz, Evgeny Stemasov, Enrico Rukzio, Teresa Hirzle
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-15472

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Put Your Muscle Into It: Introducing XEM 2 , a Novel Approach for Monitoring Exertion in Stationary Physical Exercises Leveraging Muscle Work Conference: ; ; 6166 CCS: Human-centered computing CCS: Human-centered computing Human computer interaction (HCI) CCS: Human-centered computing Interactive systems and tools Jana Franceska Funke, Mario Sagawa, Georgious Nurcan-Georgiou, Naomi Sagawa, Dennis Dietz, Evgeny Stemasov, Enrico Rukzio, Teresa Hirzle Abstract . We present a novel system for camera-based measurement and visualization of muscle work based on the Hill-Type-Muscle-Model: the exercise exertion muscle-work monitor ( XEM 2 ). Our aim is to complement and, thus, address issues of established measurement techniques that offer imprecise data for non-uniform movements (burned calories) or provide limited information on strain across different body parts (self-perception scales). We validate the reliability of XEM’s measurements through a technical evaluation of ten participants and five exercises. Further, we assess the acceptance, usefulness, benefits, and opportunities of XEM 2 in an empirical user study. Our results show that XEM 2 provides reliable values of muscle work and supports participants in understanding their workout while also providing reliable information about perceived exertion per muscle group. With this paper, we introduce a novel system capable of measuring and visualizing exertion for single muscle groups, which has the potential to improve exercise monitoring to prevent unbalanced workouts. Figure 1. The XEM 2 setup: On the left, a participant is playing an exercise game (exergame). Their movements are detected by two Azure Kinect Cameras. We then use our XEM 2 system to evaluate the exertion that different muscle groups need to perform that movement. The results are visualized in two different ways, a bar chart for prime movers (right/left upper/lower limb) and a 3D avatar with highlighted muscle groups (M. Deltoideus, M. Pectoralis Major, M. Triceps Brachii, M. Biceps Brachii and M. Brachialis, M. Latissimus Dorsi, M. Gluteus Maximus, Mm. ischiocrurales, M. Quadriceps Femoris). The picture shows a person doing exercises in front of a large screen. On the side and underneath the screen is an Azure Kinect positioned. between the person and the Kinects are dotted lines saying that the movement is detected by the Azure Kinects. From the Azure Kinect, there is a dotted line towards a box saying XEM and towards the middle avatar figure on the screen. From the XEM box there are two arrows toward the left and right side of the screen, where we can see the limb exertion boxplot and the muscle group exertion highlighted avatar. Both create a final visualization shown with a labelled arrow saying final output. The final output is the accumulated value over the whole session. 1. Introduction Monitoring physical exercise can help people achieve their fitness goals by fostering motivation. Tracking activity over time helps people set goals, allowing them to stay accountable in their fitness journey ( 26 ; 38 ; 9 ) . Besides supporting a person in following a healthy lifestyle, physical exercise monitoring permits early interventions in some cases of physical impairment (e.g., injuries), which can reduce high costs in the health sector and alleviate adverse long-term effects. Personal monitoring is especially important since physical exercise often no longer takes place in a supervised environment but increasingly alone and at home ( 11 ) . In this work, we present a novel system for exertion measurement and visualization: the Exercise Exertion Muscle-work Monitor ( XEM 2 ). XEM 2 is a camera-based movement tracking system that measures and visualizes muscle work of different muscle groups and their variation over time. The system is based on a musculoskeletal calculation using the Hill-Type Muscle Model ( 42 ) and the Kinesis Unity Plugin for calculating exertion. We use the term muscle work since we focus on the work a muscle has to accomplish to move or hold a position, though it is not the exact physical definition of work. XEM 2 provides a visualization of the used muscles both during and after an exercise has been performed. Since XEM 2 measures exertion on a muscle-specific level, it offers advantages over other exertion measurement techniques by providing detailed information. Nevertheless, we emphasize that XEM 2 was not designed to replace other measures but rather to complement them. One established way to measure physical exercise is burned calories ( 2 ) . Burned calories are calculated based on a single generic value called the metabolic equivalent (MET) , a person’s body parameters (age, weight, heart rate), and training time ( 1 ; 6 ; 2 ) . The MET is medically pre-measured in detail for each exercise type based on oxygen uptake, measured in ml/kg/min. This value is relatively accurate for representing repetitive and steady activities, where one can assume an even burn of calories, such as running, walking, and swimming. However, it is rather imprecise for exercises like yoga or dancing where movements and intensity change frequently during the exercise, but only a single approximate value is applied ( 1 ) . For exercise games (exergames), an emerging approach to home exercise motivation, there are approaches using MET values ( 17 ) . However, the associated MET value has to be pre-measured for each new game or game sequence. Secondly, there are self-perception scales like the Rating of Perceived Exertion scale (RPE) ( 36 ; 39 ) . While they provide more detailed values, they are not as accurate as objective measures, because they need practice and good self-assessment ( 15 ) . Finally, burned calories and self-perception scales only provide a single value for the exertion of an exercise. A single value gives little information about the physical strain of specific body parts and about the muscles being involved. However, more information and support for physical exercise is often desired ( 38 ; 23 ) . Our system, XEM 2 , complements these generic exercise measurement methods by providing reliable information about individual muscles. Thus, it can not only help users to have a well-balanced selection of exercises to avoid unilateral training but also prevent overload and injuries ( 28 ; 19 ; 12 ) . We measured validity and evaluated the acceptance and usefulness of XEM 2 in an empirical user study (N=36). First, we measured XEM 2 values for five different exercises and compared the change of values for different muscle groups. Secondly, we quantitatively compared participants’ acceptance of XEM 2 with two established exertion measurement methods: (1) the RPE scale and (2) burned calories. We also collected qualitative feedback from the participants about the advantages, disadvantages, and potential improvements of XEM 2 . Thirdly, we focused on the visualization of XEM 2 and qualitatively evaluated the effects and benefits of the muscle work visualized by a 3D virtual avatar and box plot. For technical validity, repetitive values showed only small fluctuations between repetitive movements but big differences between involving a muscle group or not. The findings of the user study part demonstrate that in providing useful, precise, and trustworthy information about exertion XEM 2 was about equal compared to burned calories and the RPE scale. The participants found that XEM 2 presented the most interesting and useful information about their workout. In summary, XEM 2 provides granular information about muscle work in specific muscles during

Read the complete text