xieliwei commited on
Commit
f2049b6
·
verified ·
1 Parent(s): 6b0c58c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +148 -3
README.md CHANGED
@@ -1,3 +1,148 @@
1
- ---
2
- license: cc-by-nc-sa-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-sa-4.0
3
+ task_categories:
4
+ - feature-extraction
5
+ - object-detection
6
+ language:
7
+ - en
8
+ tags:
9
+ - CSI
10
+ - WiFi
11
+ - 6E
12
+ - 6GHz
13
+ - Sensing
14
+ - timeseries
15
+ pretty_name: Wi6CSI
16
+ size_categories:
17
+ - 1K<n<10K
18
+ ---
19
+
20
+ # Wi6CSI
21
+ Wi6CSI: A Human Activity Recognition Dataset for Next-Generation WiFi Sensing in the 6 GHz Band
22
+
23
+ ## Wi6CSI Dataset Overview
24
+
25
+ **Wi6CSI** is a comprehensive Wi-Fi sensing dataset collected in the **6GHz** band. It offers a rich, high-resolution resource for researchers
26
+
27
+ ### Key Features
28
+
29
+ - **Scenarios Covered:**
30
+ - **Single User:** Data captured with one user performing various activities.
31
+ - **Multi User:** Data where multiple users are present simultaneously.
32
+ - **People Counting:** Scenarios designed to count individuals within a specific area.
33
+
34
+ - **Collection Locations:**
35
+ Data has been collected across **three distinct locations**, ensuring diverse environmental conditions and variations in the wireless channel.
36
+
37
+ - **Data Volume and Resolution:**
38
+ - The dataset holds approximately **1.5TB** of data.
39
+ - Each Wi-Fi packet includes **8100 subcarriers**, offering a level of detail that supports high-resolution sensing and analysis.
40
+
41
+ - **Wi-Fi Band:**
42
+ - Utilizes the **6GHz** frequency band, providing enhanced granularity and improved sensing accuracy compared to lower frequency bands.
43
+
44
+ ### Dataset Contents
45
+
46
+ - **.mp4 files:**
47
+ These files are Pose3D videos that capture 3D representations of subjects during data collection. They can be used for visual analysis, activity recognition, or correlating motion with CSI data.
48
+
49
+ - **.mat files:**
50
+ These MATLAB files contain the Wi-Fi CSI values. The numerical data stored in these files represent the channel state information of the wireless signal, useful for signal processing or machine learning applications.
51
+
52
+ - **.csi files:**
53
+ These files hold raw Wi-Fi CSI data. They are not immediately usable and must be parsed using the [picoscenes toolbox](https://ps.zpj.io/matlab.html) to extract the relevant features for analysis.
54
+
55
+ ### Single User Activity Recognition Structure
56
+ | **Activity** |
57
+ |---------------------------------|
58
+ | Empty Room (A1) |
59
+ | Standing Still (A2) |
60
+ | Sitting Still (A3) |
61
+ | Lying Down Still (A4) |
62
+ | Clap (A5) |
63
+ | Wave (A6) |
64
+ | Arm Circles (A7) |
65
+ | Bicep Curls (A10) |
66
+ | Toe Touches (A8) |
67
+ | Squats (A9) |
68
+ | Walk (Perpendicular) (A11) |
69
+ | Walk (Parallel) (A12) |
70
+ | Jump (A13) |
71
+
72
+ | **Location** |
73
+ |---------------------------------|
74
+ | Room (L1) |
75
+ | Extended Room (L2) |
76
+ | Meeting Room (L3) |
77
+
78
+ | **Trial Number** |
79
+ |---------------------------------|
80
+ | Trial 1 (T1) |
81
+ | Trial 2 (T2) |
82
+ | Trial 3 (T3) |
83
+ | Trial 4 (T4) |
84
+
85
+ | **Participant Number** |
86
+ |---------------------------------|
87
+ | Participant 1 (P1) |
88
+ | Participant 2 (P2) |
89
+ | Participant 3 (P3) |
90
+ | Participant 4 (P4) |
91
+ | Participant 5 (P5) |
92
+
93
+ ### Multi User Interaction Structure
94
+ | **Scenario No.** | **Participant P1** | **Participant P2** | **Participant P3** |
95
+ |------------------|------------------------|------------------------|--------------------------|
96
+ | S1 | (A5) Clap | (A6) Wave | (A7) Arm Circles |
97
+ | S2 | (A10) Bicep Curls | (A2) Standing Still | (A6) Wave |
98
+ | S3 | (A3) Sitting Still | (A10) Bicep Curls | (A6) Wave |
99
+ |------------------|------------------------|------------------------|--------------------------|
100
+ | S4 | (A13) Jump | (A12) Walk (Parallel) | (A5) Clap |
101
+ | S5 | (A10) Bicep Curls | (A9) Squats | (A12) Walk (Parallel) |
102
+ | S6 | (A9) Squats | (A6) Wave | (A2) Standing Still |
103
+ |------------------|------------------------|------------------------|--------------------------|
104
+ | S7 | (A14) High Five | (A14) High Five | (A6) Wave |
105
+ | S8 | (A15) Passing Object | (A15) Passing Object | (A12) Walk (Parallel) |
106
+ | S9 | (A16) Handshake | (A16) Handshake | (A5) Clap |
107
+
108
+ | **Location** |
109
+ |---------------------------------|
110
+ | Meeting Room (L3) |
111
+
112
+ | **Trial Number** |
113
+ |---------------------------------|
114
+ | Trial 1 (T1) |
115
+ | Trial 2 (T2) |
116
+ | Trial 3 (T3) |
117
+ | Trial 4 (T4) |
118
+
119
+ | **Participant Number** |
120
+ |---------------------------------|
121
+ | Participant 1 (P1) |
122
+ | Participant 2 (P2) |
123
+ | Participant 3 (P3) |
124
+
125
+ ### People Counting Structure
126
+ | **Scenario No.** | **Details** |
127
+ |------------------|-----------------------------------------------------------------|
128
+ | S10 | Sitting relatively still while using a mobile phone and talking |
129
+ | S11 | Standing and mingling with other participants |
130
+
131
+ | **Location** |
132
+ |---------------------------------|
133
+ | Meeting Room (L3) |
134
+
135
+ | **Trial Number** |
136
+ |---------------------------------|
137
+ | Trial 1 (T1) |
138
+ | Trial 2 (T2) |
139
+ | Trial 3 (T3) |
140
+ | Trial 4 (T4) |
141
+
142
+ | **Number of Participants** | **Participant Numbers** |
143
+ |----------------------------|-------------------------|
144
+ | 1 (N1) | P1 |
145
+ | 2 (N2) | P1, P2 |
146
+ | 3 (N3) | P1, P2, P3 |
147
+ | 4 (N4) | P1, P2, P3, P4 |
148
+ | 5 (N5) | P1, P2, P3, P4, P5 |