6
回編集
「Reliable Heading Tracking For Pedestrian Road Crossing Prediction Using Commodity Devices」の版間の差分
ナビゲーションに移動
検索に移動
細
Reliable Heading Tracking For Pedestrian Road Crossing Prediction Using Commodity Devices (編集)
2025年12月2日 (火) 21:34時点における版
、 2025年12月2日 (火) 21:34編集の要約なし
RobT468167542161 (トーク | 投稿記録) (ページの作成:「<br>Pedestrian heading tracking permits purposes in pedestrian navigation, site visitors safety, and accessibility. Previous works, utilizing inertial sensor fusion or ma…」) |
KandisGardin0 (トーク | 投稿記録) 細 |
||
| 1行目: | 1行目: | ||
<br>Pedestrian heading tracking | <br>Pedestrian heading tracking allows purposes in pedestrian navigation, traffic security, and accessibility. Previous works, utilizing inertial sensor fusion or machine studying, are limited in that they assume the phone is mounted in particular orientations, hindering their generalizability. We propose a brand new heading monitoring algorithm, the Orientation-Heading Alignment (OHA), which leverages a key insight: people tend to hold smartphones in certain methods resulting from habits, such as swinging them while walking. For each smartphone angle during this motion, OHA maps the smartphone orientation to the pedestrian heading and learns such mappings efficiently from coarse headings and smartphone orientations. To anchor our algorithm in a sensible scenario, we apply OHA to a challenging task: predicting when pedestrians are about to cross the street to enhance street person safety. Specifically, using 755 hours of walking data collected since 2020 from 60 people, we develop a lightweight mannequin that operates in real-time on commodity units to foretell street crossings. Our analysis reveals that OHA achieves 3.4 occasions smaller heading errors across nine situations than present methods.<br><br><br><br>Furthermore, OHA allows the early and accurate detection of pedestrian crossing behavior, issuing crossing alerts 0.35 seconds, on average, earlier than pedestrians enter the road vary. Tracking pedestrian heading involves constantly monitoring an individual’s going through course on a 2-D flat plane, typically the horizontal plane of the worldwide coordinate system (GCS). Zhou et al., 2014). For example, a pedestrian could possibly be strolling from south to north on a highway while swinging a smartphone. In this case, smartphone orientation estimation would indicate the device’s dynamic orientation relative to the GCS, commonly represented by Euler angles (roll, pitch, yaw). However, tracking pedestrian heading should precisely show that the pedestrian is moving from south to north, regardless of how the smartphone is oriented. Existing approaches to estimating pedestrian heading by IMU (Inertial Measurement Unit) make use of a two-stage pipeline: first, they estimate the horizontal aircraft utilizing gravity or magnetic fields, after which integrate the gyroscope to trace relative heading adjustments (Manos et al., 2018; Thio et al., 2021; Deng et al., 2015). These approaches hinge on a crucial assumption: the telephone must stay static relative to the pedestrian physique.<br><br><br><br>We suggest a new heading monitoring algorithm, Orientation-Heading Alignment (OHA), which leverages a key insight: folks tend to hold smartphones in certain attitudes attributable to habits, whether or not swinging them while strolling, stashing them in pockets, or inserting them in luggage. These attitudes or relative orientations, outlined as the smartphone’s orientation relative to the human body quite than GCS, mainly depend on the user’s habits, traits, or even clothing. As an illustration, regardless of which course a pedestrian faces, they swing the smartphone in their habitual method. For every smartphone attitude, OHA maps the smartphone orientation to the pedestrian heading. Because the attitudes are comparatively stable for every particular person (e.g., holding a smartphone in the precise hand and swinging), it is feasible to study the mappings effectively from coarse headings and smartphone orientation. Previous analysis (Liu et al., 2023; Yang et al., 2020; Lee et al., 2023) has noted an identical insight however adopted a special approach for heading monitoring: amassing IMU and accurate heading data for multiple smartphone attitudes and coaching a machine studying mannequin to predict the heading.<br><br><br><br>However, resulting from machine discrepancies and varying consumer behaviors, it's not feasible to construct a machine studying mannequin that generalizes to all attainable smartphone attitudes. To anchor our heading estimation algorithm in a sensible scenario, we apply OHA to a challenging task: predicting when pedestrians are about to cross the road-an necessary drawback for enhancing road person safety (T., pril; Zhang et al., 2021, 2020). This activity, which requires accurate and well timed predictions of pedestrian crossings, [https://debunkingnase.org/index.php/GPS_E-Lock_For_Containers_Tanker_Trucks_Consignment_Tracking_Device iTagPro product] is further complicated by the numerous crossing patterns of pedestrians and the complexity of street layouts. Based on the OHA heading, we suggest PedHat, a lightweight, infrastructure-free system that predicts when a pedestrian is about to cross the closest road and points crossing alerts. PedHat incorporates a lightweight mannequin that accepts OHA headings as inputs and operates in real-time on user units to predict road crossings. We developed this mannequin using data we collected since 2020 from 60 people, [https://ykm.de/itagpro-87427 iTagPro product] every contributing two months of traces, protecting 755 hours of strolling data.<br> | ||