In-Bed Posture Classification Based on Sparse Representation in Redundant Dictionaries

Mihálik, Ondrej; Sýkora, T.; Hušák, M.; Fiedler, P.

doi:10.1016/j.ifacol.2022.06.062

Cited by 8 publications

(4 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• on the stomach -class 4. The dataset is similar to the set presented in [7], yet larger. It comprises 18 different test subjects in varying positions, containing 290 pressure map images in total.…”

Section: Sparse Representation Of Measurementsmentioning

confidence: 99%

“…Training images are reshaped into column vectors, normalized to unit ℓ 2 norm, and stacked into a new matrix so as to form the redundant dictionary X. The dictionary is appended by an identity matrix, which affords and effective means of dealing with occlusions [3], [7].…”

Section: Sparse Representation Of Measurementsmentioning

confidence: 99%

“…We will consider the fault being missing rows as it is a consequence of a nonfunctional sensor. To make classification possible despite errors in the data, new elements are appended to the dictionary X representing potential errors [7]. However the fault definition is interchangeable if a user has more information about typical fault character in specific case.…”

Section: Occlusionsmentioning

confidence: 99%

See 2 more Smart Citations

Sparse Representation for Classification of Posture in Bed

Mesárošová,

Mihálik

2023

Proceedings II of the 29st Conference STUDENT EEICT 2023: Selected Papers.

View full text Add to dashboard Cite

Redundant dictionaries, also known as frames, offer a non-orthogonal representation of signals, which leads to sparsity in their representative coefficients. As this approach provides many advantageous properties it has been used in various applications such as denoising, robust transmissions, segmentation, quantum theory and others. This paper investigates the possibility of using sparse representation in classification, comparing the achieved results to other commonly used classifiers. The different methods were evaluated in a real-world classification task in which the position of a lying patient has to be deduced based on the data provided by a pressure mattress of 30×11 sensors. The investigated method outperformed most of the commonly used classifiers with accuracy exceeding 92%, while being less demanding on design and implementation complexity.

show abstract

Section: Sparse Representation Of Measurementsmentioning

confidence: 99%

Section: Sparse Representation Of Measurementsmentioning

confidence: 99%

Section: Occlusionsmentioning

confidence: 99%

See 1 more Smart Citation

Sparse Representation for Classification of Posture in Bed

Mesárošová,

Mihálik

2023

Proceedings II of the 29st Conference STUDENT EEICT 2023: Selected Papers.

View full text Add to dashboard Cite

show abstract

“…In-Bed Posture Classification Based on Sparse Representation in Redundant Dictionaries [18] used pressure maps for posture classification without using other parameters such as height or weight, in an attempt to find the differences between using the classification systems known as sparse representation classification and standard classification methods.…”

Section: Analysis Of Included Studiesmentioning

confidence: 99%

Lying-People Pressure-Map Datasets: A Systematic Review

Fonseca

Ribeiro

Metrôlho

2022

Data

View full text Add to dashboard Cite

Bedded or lying-people pressure-map datasets can be used to identify patients’ in-bed postures and can be very useful in numerous healthcare applications. However, the construction of these datasets is not always easy, and many researchers often resort to existing datasets to carry out their experiments and validate their solutions. This systematic review aimed to identify and characterise pressure-map datasets on lying-people- or bedded-people positions. We used a systematic approach to select nine studies that were thoroughly reviewed and summarised them considering methods of data collection, fields considered in the datasets, and results or their uses after collection. As a result of the review, six research questions were answered that allowed a characterisation of existing datasets regarding of the types of data included, number and types of poses considered, participant characteristics and size of the dataset, and information on how the datasets were built. This study might represent an important basis for academics and researchers to understand the information collected in each pressure-map dataset, the possible uses of such datasets, or methods to build new datasets.

show abstract