A review of vision-based indoor HAR: state-of-the-art, challenges, and future prospects

With the advent of technology, we are getting more comfortable with the use of gadgets, cameras, etc., and find Artificial Intelligence as an integral part of most of the tasks we perform throughout the day. In such a scenario, the use of cameras and vision-based sensors comes as an escape from many real-time problems and challenges. One major application of these vision-based systems is Indoor Human Activity Recognition (HAR) which serves in a variety of scenarios ranging from smart homes, elderly care, assisted living, and human behavior pattern analysis for identifying any abnormal behavior to abnormal activity recognition like falling, slipping, domestic violence, etc. The effect of HAR in real time has made the area of indoor activity recognition a more explored zone by the industrial segment to attract users with their products in multiple domains. Hence, considering these aspects of HAR, this work proposes a detailed survey on indoor HAR. Through this work, we have highlighted the recent methodologies and their performance in the field of indoor activity recognition. We have also discussed- the challenges, detailed study of approaches with real-world applications of indoor-HAR, datasets available for indoor activity, and their technical details in this work. We have proposed a taxonomy for indoor HAR and highlighted the state-of-the-art and future prospects by mentioning the research gaps and the shortcomings of recent surveys with respect to our work.

Effect of balance training using virtual reality-based serious games in individuals with total knee replacement: A randomized controlled trial

BACKGROUND

Virtual reality (VR) and serious games (SGs) are widespread in rehabilitation for many orthopedic and neurological diseases. However, few studies have addressed the effects of rehabilitation with VR-based SGs on clinical, gait, and postural outcomes in individuals with total knee replacement (TKR).

OBJECTIVE

The primary objective was the efficacy of balance training using non-immersive VR-based SGs compared to conventional therapy in TKR patients on the Time Up and Go test. Secondary objectives included the efficacy on clinical, gait, and postural outcomes.

METHODS

We randomly allocated 56 individuals with unilateral TKR to the experimental group (EG) or control group (CG) for 15 sessions (45 min; 5 times per week) of non-immersive VR-based SGs or conventional balance training, respectively. The primary outcome was functional mobility measured by the Timed Up and Go test; secondary outcomes were walking speed, pain intensity, lower-limb muscular strength, independence in activities of daily living as well as gait and postural parameters.

RESULTS

We found significant within-group differences in all clinical outcomes and in a subset of gait (p<0.0001) and postural (p ≤ 0.05) parameters. Analysis of the stance time of the affected limb revealed significant between-group differences (p = 0.022): post-hoc analysis revealed within-group differences in the EG (p = 0.002) but not CG (p = 0.834). We found no significant between-group differences in other outcomes.

CONCLUSIONS

Balance training with non-immersive VR-based SGs can improve clinical, gait, and postural outcomes in TKR patients. It was not superior to the CG findings but could be considered an alternative to the conventional approach and can be added to a regular rehabilitation program in TKR patients. The EG had a more physiological duration of the gait stance phase at the end of the treatment than the CG.

CLINICALTRIALS

GOV: NCT03454256.

Overground wearable powered exoskeleton for gait training in subacute stroke subjects: clinical and gait assessments

BACKGROUND

Wearable powered exoskeletons provide intensive overground gait training with patient's active participation: these features promote a successful active motor relearning of ambulation in stroke survivors.

AIM

The aim of this study was to investigate the feasibility and the clinical effects of an overground exoskeleton-assisted gait training (OEAGT) in subacute stroke patients.

DESIGN

Prospective, pilot pre-post, open label, non-randomized experimental study.

SETTING

Four Italian neurological rehabilitation centers.

POPULATION

Forty-eight subacute stroke patients were enrolled. Two patients dropped out because of medical problems. Data analysis was conducted on 46 subjects (56.84±14.29 years; 27 male; 29 ischemic; 24 left hemiparesis).

METHODS

Patients underwent 15±2 sessions (60 min/session, 3-5 times/week) of OEAGT. Clinical and gait assessments were performed at the beginning (T1) and at the end (T2) of the training period: modified Barthel Index (BI), modified Ashworth Scale at Hip (MAS-H), Knee (MAS-K), and Ankle (MAS-A) level, Motricity Index (MI), Trunk Control Test (TCT), Functional Ambulation Classification (FAC), Walking Handicap Scale (WHS), 10-Meter Walking Test (10MWT), 6-Minute Walking Test (6mWT), Timed Up-and-Go test (TUG). The Technology Acceptance Model (TAM) questionnaire evaluated the acceptance of OEAGT by patients. Data stratification was performed using the time post the acute event and the onset of rehabilitation treatment, and the MI at T1. Wilcoxon's test (P<0.05) was used.

RESULTS

All clinical scales significantly improved at T2; no statistically significant changes were reported for MAS-H, MAS-K, MAS-A. The 69.57% patients were able to walk at T1; 17.39% were not able to walk at T1 but regained ambulation at T2; and 13.04% were not able to walk at either T1 or T2. The ambulant patients showed a statistical improvement in speed measured during the 10MWT and in the distance covered over a time of 6 minutes (6mWT). The results from the TAM questionnaire showed that all subjects perceived the OEAGT positively. The data stratification analysis suggests that the OEAGT does not have any restriction of use.

CONCLUSIONS

The OEAGT improved the clinical and gait outcomes in subacute patients. Randomized studies on larger samples are needed to confirm these data and to assess the efficacy of OEAGT.

CLINICAL REHABILITATION IMPACT

Introduce innovative rehabilitation strategies based on customized OEAGT.

Using a marker-less method for estimating L5/S1 moments during symmetrical lifting

The aim of this study is to analyze the validity of a computer vision-based method to estimate 3D L5/S1 joint moment during symmetrical lifting. An important criterion to identify the non-ergonomic lifting task is the value of net moment at L5/S1 joint. This is usually calculated in a laboratory environment which is not practical for on-site biomechanical analysis. The validity of the proposed method, was assessed externally by comparing the results with a lab-based reference method and internally by comparing the estimated L5/S1 joint moments from top-down model and bottom-up model. It was shown that no significant differences in peak and mean moments between the two methods and intra-class correlation coefficients revealed excellent reliability of the proposed method (>0.91). The proposed method provides a reliable tool for assessment of lower back loads during occupational lifting and can be an alternative when the use of marker-based motion tracking systems is not possible.

Linear discriminant analysis for symmetric lifting recognition of skilled logistic experts by center of pressure trajectory

GOAL

The main purpose of the present study was to propose a recognizing method to analyze characteristics of symmetric lifting of skilled logistic experts by center of pressure (CoP) trajectories. Although it has been known that good posture helps reduce the intradiscal loads on lumbar discs, the most significant problem was that most of logistic workers did not know whether the current posture was proper or not.

METHODS

The experiment of lifting was performed three times under 18 kg loads with closed eyes. Six skilled logistic experts and six unskilled beginners were participated in. The linear discriminant analysis (LDA) was designed by seven indices which were derived from measured CoP trajectories with the Wii Balance Board. The strong point of experimental system was practical, reliable, and cheap.

RESULTS

As a result, it was found that the designed LDA discriminated difference of symmetric lifting between skilled experts and unskilled beginners with the error rate of 0.005.

DISCUSSION

It was discussed that not only most of unskilled beginners had mainly characteristics of poor lifting posture, but also the proposed method showed a high possibility to self-evaluate the symmetric lifting in order to check whether the current posture of logistic worker is proper or not.

Use of the sit-to-stand task to evaluate motor function of older adults using telemetry

Background

Physical exercises are widely used in community programs, but not all older adults are willing to participate. Information and communication technology may solve this problem by allowing older people to participate in fitness programs at home. Use of remote instruction will facilitate physical exercise classes without requiring that participants gather at one place. The aim of this study was to examine use of a sit-to-stand task in evaluating motor function using conventional video communication in a telemetry system to enable real-time monitoring, and evaluation in physical performance of older adults at home.

Methods

The participants were 59 older individuals and 81 university students. Three physical exercise batteries were used: arm curl, figure-of-eight walk test, and functional reach. The knee extension maximum angular velocity (KEMAV) and the iliac elevation maximum velocity (IEMV) during standing up from a chair and the heel rise frequency were used in the motion-capture measurements. The results were assessed using multi-group structural equation modeling (SEM) for the young and older groups.

Results

Young participants consistently performed better than their older counterparts on all items. Analyses with multi-group SEM based on correlations between items yielded a good model-fit for the data. Among all path diagrams for IEMV and KEMAV in the older and young groups, paths from muscular strength to skillfulness showed significant effects. The path from the IEMV to muscular strength was also significant in the older group.

Conclusions

Multi-group SEM suggested that video-based measurements of IEMV during sit-to-stand motion can estimate muscular strength, which suggests that remote monitoring of physical performance can support wellness of community-dwelling older adults.

Development of a bio-inspired mechatronic chest wall simulator for evaluating the performances of opto-electronic plethysmography

Instrumented gait analysis based on optoelectronic systems is an expensive technique used to objectively measure the human movement features and it is generally considered as the gold standard. Opto-electronic plethysmography (OEP) is a particular motion analysis system able to: (i) determine chest wall kinematic via the evaluation of marker displacements placed on the thorax and (ii) compute respiratory volumes during breathing.

The aim of this work is to describe the performances of a custom made, bio-inspired, mechatronic chest wall simulator (CWS), specifically designed to assess the metrological performances of the OEP system. The design of the simulator is based on the chest wall kinematic analysis of three healthy subjects previously determined.

Two sets of experiments were carried out: (i) to investigate the CWS dynamic response using different target displacements (1 - 12 mm), and (ii) to assess the CWS accuracy and precision in simulating quite breathing, covering the physiological range of respiratory frequency and tidal volume.

Results show that the CWS allows simulating respiratory frequency up to ~ 60 bpm. The difference between the actual displacement and the set one is always < 9 μm. The precision error, expressed as the ratio between measurement uncertainty and the actual displacement, is lower than 0.32 %.

The observed good performances permit to consider the CWS prototype feasible to be employed for assessing the performances of OEP system in periodical validation routines.

Significant body point labeling and tracking

In this paper, a method is presented to label and track anatomical landmarks (e.g., head, hand/arm, feet), which are referred to as significant body points (SBPs), using implicit body models. By considering the human body as an inverted pendulum model, ellipse fitting and contour moments are applied to classify it as being in Stand, Sit, or Lie posture. A convex hull of the silhouette contour is used to determine the locations of SBPs. The particle filter or a motion flow-based method is used to predict SBPs in occlusion. Stick figures of various activities are generated by connecting the SBPs. The qualitative and quantitative evaluation show that the proposed method robustly labels and tracks SBPs in various activities of two different (low and high) resolution data sets.