1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
| """
触觉警告模态注意力转移序列分析
基于 MDPI Systems 14(3):328 (2026)
核心方法：
1. 记录警告前后的注视点序列
2. 将注视分类为：道路/仪表/中控/手机/其他
3. 分析警告后的注视转移模式
4. 序列分析评估注意力回归速度
关键发现：
- 触觉警告 → 道路注视恢复最快（0.6s）
- 听觉警告 → 检测率最高但用户体验差
- 视觉警告 → 最慢（1.2s），分心时可能看不到
- 触觉+听觉 → 最佳综合表现
"""
import numpy as np
from typing import List, Tuple, Optional
from dataclasses import dataclass
@dataclass
class AlertModalityResult:
"""警告模态实验结果"""
modality: str
detection_rate: float
response_time_ms: int
road_fixation_recovery: float
user_preference: float
false_alarm_rate: float
class AttentionTransitionAnalyzer:
"""
注意力转移序列分析器
将注视点序列编码为状态序列，
分析警告后的状态转移模式
状态定义：
R = Road（道路）
D = Dashboard（仪表）
C = Center console（中控）
P = Phone（手机）
O = Other（其他）
"""
STATES = ['R', 'D', 'C', 'P', 'O']
MODALITY_RESULTS = {
'visual_only': AlertModalityResult(
modality='visual_only',
detection_rate=85.0,
response_time_ms=1200,
road_fixation_recovery=0.72,
user_preference=0.55,
false_alarm_rate=0.08
),
'auditory_only': AlertModalityResult(
modality='auditory_only',
detection_rate=95.0,
response_time_ms=800,
road_fixation_recovery=0.85,
user_preference=0.40,
false_alarm_rate=0.12
),
'haptic_only': AlertModalityResult(
modality='haptic_only',
detection_rate=90.0,
response_time_ms=600,
road_fixation_recovery=0.91,
user_preference=0.75,
false_alarm_rate=0.05
),
'haptic_auditory': AlertModalityResult(
modality='haptic_auditory',
detection_rate=98.0,
response_time_ms=500,
road_fixation_recovery=0.96,
user_preference=0.60,
false_alarm_rate=0.06
),
}
def analyze_transition_sequence(self, gaze_sequence: List[str],
alert_time: int) -> dict:
"""
分析警告前后的注视转移序列
Args:
gaze_sequence: 注视状态序列 ['R', 'P', 'C', 'R', ...]
alert_time: 警告触发时刻（序列索引）
Returns:
analysis: 转移模式分析
"""
pre_alert = gaze_sequence[:alert_time]
post_alert = gaze_sequence[alert_time:]
pre_road_ratio = sum(1 for s in pre_alert if s == 'R') / max(len(pre_alert), 1)
recovery_time = 0
for i, state in enumerate(post_alert):
if state == 'R':
recovery_time = i
break
post_10 = post_alert[:10]
post_road_ratio = sum(1 for s in post_10 if s == 'R') / max(len(post_10), 1)
transitions = {}
for i in range(len(gaze_sequence) - 1):
pair = (gaze_sequence[i], gaze_sequence[i+1])
transitions[pair] = transitions.get(pair, 0) + 1
road_return_transitions = {
k: v for k, v in transitions.items()
if k[1] == 'R' and k[0] != 'R'
}
return {
'pre_alert_road_ratio': pre_road_ratio,
'post_alert_road_ratio': post_road_ratio,
'recovery_time_frames': recovery_time,
'recovery_time_ms': recovery_time * 100,
'road_return_transitions': road_return_transitions,
'improvement': post_road_ratio - pre_road_ratio,
}
def compare_modalities(self) -> dict:
"""对比各警告模态"""
results = {}
for modality, result in self.MODALITY_RESULTS.items():
results[modality] = {
'detection_rate': result.detection_rate,
'response_time_ms': result.response_time_ms,
'road_recovery': result.road_fixation_recovery,
'user_preference': result.user_preference,
'false_alarm_rate': result.false_alarm_rate,
}
return results
def recommend_modality(self, scenario: str) -> str:
"""
根据场景推荐最佳警告模态
Args:
scenario: 'visual_distraction'/'cognitive_distraction'/'fatigue'/'emergency'
"""
recommendations = {
'visual_distraction': 'haptic_only',
'cognitive_distraction': 'haptic_auditory',
'fatigue': 'haptic_auditory',
'emergency': 'haptic_auditory',
'normal': 'haptic_only',
}
return recommendations.get(scenario, 'haptic_only')
if __name__ == "__main__":
analyzer = AttentionTransitionAnalyzer()
print("=== 警告模态对比 ===")
results = analyzer.compare_modalities()
print(f"{'模态':<20} {'检测率%':<10} {'响应ms':<10} {'道路恢复%':<12} {'偏好':<8} {'误报%':<8}")
print("-" * 70)
for modality, data in results.items():
print(f"{modality:<20} {data['detection_rate']:<10.1f} "
f"{data['response_time_ms']:<10} {data['road_recovery']:<12.2f} "
f"{data['user_preference']:<8.2f} {data['false_alarm_rate']:<8.2f}")
np.random.seed(42)
distracted_gaze = ['R', 'R', 'P', 'P', 'P', 'C', 'P', 'P',
'R', 'R', 'R', 'R', 'R']
analysis = analyzer.analyze_transition_sequence(distracted_gaze, alert_time=8)
print(f"\n=== 模拟分心→警告→恢复 ===")
print(f"警告前道路注视: {analysis['pre_alert_road_ratio']:.2f}")
print(f"警告后道路注视(10帧): {analysis['post_alert_road_ratio']:.2f}")
print(f"恢复时间: {analysis['recovery_time_ms']}ms")
print(f"改善: {analysis['improvement']:.2f}")
print(f"\n=== 场景推荐 ===")
for scenario in ['visual_distraction', 'cognitive_distraction',
'fatigue', 'emergency', 'normal']:
rec = analyzer.recommend_modality(scenario)
print(f" {scenario}: {rec}")
print(f"\n=== 触觉反馈技术实现 ===")
print(f" 方向盘振动电机: LRA线性谐振执行器")
print(f" 振动频率: 150-250 Hz（最佳皮肤感知）")
print(f" 振动模式: 单脉冲/双脉冲/连续/渐强")
print(f" 定位精度: 左右分离（左右手感知不同方向）")
print(f" 延迟: <50ms（从触发到振动）")
print(f" 成本: $3-8/方向盘")
|
