Datasets:

Modalities:
Image
HapticSensingLab commited on
Commit
c50e957
·
verified ·
1 Parent(s): 9105f12

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +144 -3
README.md CHANGED
@@ -1,3 +1,144 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # EventTouch Dataset
2
+
3
+ ## 1. Overview
4
+
5
+ The EventTouch dataset contains data from tactile-sensor contact experiments and finite element method (FEM) simulations. It includes sensor outputs, contact forces, nodal displacements, and experimental parameters collected for different objects, contact poses, and press depths.
6
+
7
+ The dataset can be used for:
8
+
9
+ - tactile perception research;
10
+ - contact-force prediction;
11
+ - FEM simulation validation;
12
+ - comparison of simulated and real-world data; and
13
+ - tactile-sensor calibration.
14
+
15
+ ## 2. Dataset Structure
16
+
17
+ ```text
18
+ data/
19
+ └── sensor/
20
+ └── object/
21
+ └── pose/
22
+ └── press_depth/
23
+ ├── fem_sensor.csv
24
+ ├── fem_node.csv
25
+ ├── real_sensor.csv
26
+ └── camera_frames/
27
+ ```
28
+
29
+ Directory descriptions:
30
+ - `sensor` identifies different types of tactile sensor.
31
+ - `object` identifies the geometry, material, and radius of the contact object(mm).
32
+ - `pose` identifies the contact pose of the object.
33
+ - `press_depth` identifies the press depth.
34
+ - Each lowest-level directory contains the data from one complete experiment.
35
+
36
+ ## 3. File Descriptions
37
+
38
+ ### `fem_sensor.csv`
39
+
40
+ Contains the simulated six-axis force/torque data. Each row corresponds to one simulation frame. The number of data rows, excluding the header, is the number of valid frames in the experiment.
41
+
42
+ ### `fem_node.csv`
43
+
44
+ Contains the displacements and contact forces of the tactile sensor's surface nodes.
45
+
46
+ ### `real_sensor.csv`
47
+
48
+ Contains the raw and processed measurements from the physical sensor experiment.
49
+
50
+ ### `camera_frames`
51
+
52
+ A folder contains camera frames of vision_based tactile sensor.
53
+
54
+ ## 4. Field Descriptions
55
+
56
+ ### Sensor types
57
+ - `digit`: digit sensor;
58
+ - `s1`: sphere vision-based tactile sensor;
59
+ - `s2`: ellipsoid vision-based tactile sensor;
60
+ - `s3`: large area resister-based tactle sensor;
61
+
62
+
63
+ ### Contact poses
64
+
65
+ For spherical objects:
66
+
67
+ - `pose1`: centered pressing;
68
+ - `pose2`: gripper offset upward by 2 mm;
69
+ - `pose3`: gripper offset upward by 4 mm;
70
+ - `pose4`: gripper offset downward by 2 mm; and
71
+ - `pose5`: gripper offset downward by 4 mm.
72
+
73
+ For cylindrical objects:
74
+
75
+ - `pose1`: centered pressing;
76
+ - `pose2`: gripper offset forward by 2 mm;
77
+ - `pose3`: gripper offset forward by 4 mm;
78
+ - `pose4`: gripper offset backward by 2 mm; and
79
+ - `pose5`: gripper offset backward by 4 mm.
80
+
81
+ ### Material codes
82
+
83
+ | Code | Material |
84
+ |---|---|
85
+ | `0010` | Ecoflex 00-10 |
86
+ | `0030` | Ecoflex 00-30 |
87
+ | `0050` | Ecoflex 00-50 |
88
+ | `15A` | Dragon Skin 15 |
89
+ | `PLA` | PLA |
90
+
91
+ ### `fem_sensor.csv`
92
+
93
+ | Field | Description | Unit |
94
+ |---|---|---|
95
+ | `frame` | Frame index, starting from 0 | Dimensionless |
96
+ | `Fx` | Force along the x-axis | N |
97
+ | `Fy` | Force along the y-axis | N |
98
+ | `Fz` | Force along the z-axis | N |
99
+ | `Mx` | Moment about the x-axis | N·m |
100
+ | `My` | Moment about the y-axis | N·m |
101
+ | `Mz` | Moment about the z-axis | N·m |
102
+
103
+ ### `fem_node.csv`
104
+
105
+ | Field | Description | Unit |
106
+ |---|---|---|
107
+ | `frame` | Frame index, starting from 0 | Dimensionless |
108
+ | `time_s` | Simulation time | s |
109
+ | `node_id` | Node identifier | Dimensionless |
110
+ | `x_m` | x-coordinate of the node | m |
111
+ | `y_m` | y-coordinate of the node | m |
112
+ | `z_m` | z-coordinate of the node | m |
113
+ | `ux_m` | Node displacement along the x-axis | m |
114
+ | `uy_m` | Node displacement along the y-axis | m |
115
+ | `uz_m` | Node displacement along the z-axis | m |
116
+ | `fx_N` | Contact force on the node along the x-axis | N |
117
+ | `fy_N` | Contact force on the node along the y-axis | N |
118
+ | `fz_N` | Contact force on the node along the z-axis | N |
119
+
120
+ ### `real_sensor.csv`
121
+
122
+ | Field | Description | Unit |
123
+ |---|---|---|
124
+ | `frame` | Frame index, starting from 0 | Dimensionless |
125
+ | `sensor_ros_time_ns` | Sensor timestamp | ns |
126
+ | `phase` | Gripper operating phase | Dimensionless |
127
+ | `device_raw_fx` | Raw x-axis force measured by the six-axis force/torque sensor | N |
128
+ | `device_raw_fy` | Raw y-axis force measured by the six-axis force/torque sensor | N |
129
+ | `device_raw_fz` | Raw z-axis force measured by the six-axis force/torque sensor | N |
130
+ | `device_raw_mx` | Raw x-axis moment measured by the six-axis force/torque sensor | N·m |
131
+ | `device_raw_my` | Raw y-axis moment measured by the six-axis force/torque sensor | N·m |
132
+ | `device_raw_mz` | Raw z-axis moment measured by the six-axis force/torque sensor | N·m |
133
+ | `corrected_fx_F_N` | Zero-offset-corrected x-axis force | N |
134
+ | `corrected_fy_F_N` | Zero-offset-corrected y-axis force | N |
135
+ | `corrected_fz_F_N` | Zero-offset-corrected z-axis force | N |
136
+ | `corrected_mx_F_Nm` | Zero-offset-corrected x-axis moment | N·m |
137
+ | `corrected_my_F_Nm` | Zero-offset-corrected y-axis moment | N·m |
138
+ | `corrected_mz_F_Nm` | Zero-offset-corrected z-axis moment | N·m |
139
+ | `camera_image_path` | Path to the visual-tactile sensor image associated with the current frame | N/A |
140
+
141
+ ## 5. Notes
142
+
143
+ 1. The raw 1,000 Hz signals from the physical six-axis force/torque sensor are downsampled to 60 Hz.
144
+ 2. Images from the visual-tactile sensor are captured at 30 frames per second.