No abstract

No abstract

The application of deep learning techniques resulted in remarkable improvement of machine learning models. In this paper we provide detailed characterizations of deep learning models used in many Facebook social network services. We present computational characteristics of our models, describe high-performance optimizations targeting existing systems, point out their limitations and make suggestions for the future general-purpose/accelerated inference hardware. Also, we highlight the need for better co-design of algorithms, numerics and computing platforms to address the challenges of workloads often run in data centers.

BackgroundAnalysis of gait features provides important information during the treatment of neurological disorders, including Parkinson’s disease. It is also used to observe the effects of medication and rehabilitation. The methodology presented in this paper enables the detection of selected gait attributes by Microsoft (MS) Kinect image and depth sensors to track movements in three-dimensional space.Methods The experimental part of the paper is devoted to the study of three sets of individuals: 18 patients with Parkinson’s disease, 18 healthy aged-matched individuals, and 15 students. The methodological part of the paper includes the use of digital signal-processing methods for rejecting gross data-acquisition errors, segmenting video frames, and extracting gait features. The proposed algorithm describes methods for estimating the leg length, normalised average stride length (SL), and gait velocity (GV) of the individuals in the given sets using MS Kinect data.ResultsThe main objective of this work involves the recognition of selected gait disorders in both the clinical and everyday settings. The results obtained include an evaluation of leg lengths, with a mean difference of 0.004 m in the complete set of 51 individuals studied, and of the gait features of patients with Parkinson’s disease (SL: 0.38 m, GV: 0.61 m/s) and an age-matched reference set (SL: 0.54 m, GV: 0.81 m/s). Combining both features allowed for the use of neural networks to classify and evaluate the selectivity, specificity, and accuracy. The achieved accuracy was 97.2 %, which suggests the potential use of MS Kinect image and depth sensors for these applications.ConclusionsDiscussion points include the possibility of using the MS Kinect sensors as inexpensive replacements for complex multi-camera systems and treadmill walking in gait-feature detection for the recognition of selected gait disorders.

This paper is devoted to a new method of using Microsoft (MS) Kinect sensors for non-contact monitoring of breathing and heart rate estimation to detect possible medical and neurological disorders. Video sequences of facial features and thorax movements are recorded by MS Kinect image, depth and infrared sensors to enable their time analysis in selected regions of interest. The proposed methodology includes the use of computational methods and functional transforms for data selection, as well as their denoising, spectral analysis and visualization, in order to determine specific biomedical features. The results that were obtained verify the correspondence between the evaluation of the breathing frequency that was obtained from the image and infrared data of the mouth area and from the thorax movement that was recorded by the depth sensor. Spectral analysis of the time evolution of the mouth area video frames was also used for heart rate estimation. Results estimated from the image and infrared data of the mouth area were compared with those obtained by contact measurements by Garmin sensors (www.garmin.com). The study proves that simple image and depth sensors can be used to efficiently record biomedical multidimensional data with sufficient accuracy to detect selected biomedical features using specific methods of computational intelligence. The achieved accuracy for non-contact detection of breathing rate was 0.26% and the accuracy of heart rate estimation was 1.47% for the infrared sensor. The following results show how video frames with depth data can be used to differentiate different kinds of breathing. The proposed method enables us to obtain and analyse data for diagnostic purposes in the home environment or during physical activities, enabling efficient human-machine interaction.