2020
DOI: 10.1109/access.2020.3006067
|View full text |Cite
|
Sign up to set email alerts
|

Depth Sequential Information Entropy Maps and Multi-Label Subspace Learning for Human Action Recognition

Abstract: Human action recognition plays a key role in human-computer interaction in complex environments. However, similar actions will lead to poor feature sequence extraction and result in a reduction in recognition accuracy. This paper proposes a method (Action-Fusion: Multi-label subspace Learning (MLSL)) from depth maps called Depth Sequential Information Entropy Maps (DSIEM) and skeleton data for human action recognition in multiple modal features. The DSIEM describe the spatial information of human motion with i… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
11
0
3

Year Published

2020
2020
2024
2024

Publication Types

Select...
4
2
2

Relationship

1
7

Authors

Journals

citations
Cited by 25 publications
(14 citation statements)
references
References 41 publications
0
11
0
3
Order By: Relevance
“…According to the type of input data, the human action recognition technologies consist of RGB video based methods [7][8][9], depth video based methods [16,18,56,60], depth skeleton based methods [6,36,45], and multi-modal data fusion based methods [54,55,57]. Due to the convenience of data acquisition and invariance to illumination and texture changes, many researchers focus on depth video based methods.…”
Section: Related Workmentioning
confidence: 99%
“…According to the type of input data, the human action recognition technologies consist of RGB video based methods [7][8][9], depth video based methods [16,18,56,60], depth skeleton based methods [6,36,45], and multi-modal data fusion based methods [54,55,57]. Due to the convenience of data acquisition and invariance to illumination and texture changes, many researchers focus on depth video based methods.…”
Section: Related Workmentioning
confidence: 99%
“…LHS RHS підтримка довіра піднесення номер [1] {Охорона здоров'я} => {Фінанси} 0.007692308 0.09090909 0.5371901 1 [2] {Фінанси} => {Охорона здоров'я} 0.007692308 0.04545455 0.5371901 1 [3] {Спорт} => {Розваги} 0.015384615 0.14285714 1.1607143 2 [4] {Розваги} => {Спорт} 0.01538461 0.12500000 1.1607143 2 [5] {Спорт} => {Фінанси} 0.015384615 0.14285714 0.8441558 2 [6] {Фінанси} => {Спорт} 0.015384615 0.09090909 0.8441558 2 [7] {Забави} => {Фінанси} 0.007692308 0.06250000 0.3693182 1 [8] {Фінанси} => {Розваги} 0.007692308 0.04545455 0.3693182 1 [9] {Покупки} => {Водіння} 0.046153846 0.46153846 1.8181818 6 [10] {Водіння} => {Покупки} 0.046153846 0.18181818 1.8181818 6 [11] {Відпочинок} => {Фінанси} 0.015384615 0.11111111 0.6565657 2 [12] {Фінанси} => {Відпочинок} 0.015384615 0.09090909 0.6565657 2 Отже, правила 3, 4, 9, 10 важливі для аналізу. Далі знайдено підтримку наступних правил (рис.…”
Section: табл 2 параметри асоціативних правилunclassified
“…Порівняємо наші результати з відомими методами. У роботах [5,12] використовують двоступеневу модель для розпізнавання образів людини. Перша частина мо-делі, заснована на розширенні ConvNets до 3D-випадку, автоматично вивчає просторово-часові особливості.…”
Section: табл 5 матриця невідповідностейunclassified
See 1 more Smart Citation
“…Chao et al [14] proposed depth spatial-temporal energy map (DSTEM) for representing the temporal information, energy information, and spatial structure information of actions. Yang et al [15] proposed a method by Multi-label subspace Learning (MLSL) to fuse depth sequential information entropy map (DSIEM) and skeleton data. Chao et al [16] proposed a motion collaborative Spatiotemporal vector (MCSTV), which comprehensively considers the integral and cooperative between the moving joints of the human body.…”
Section: Introductionmentioning
confidence: 99%