English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The resolution of GPS measurements, especially in urban areas, isinsufficient for identifying a vehicle's lane. In this work, we develop a deepLSTM neural network model LaNet that determines the lane vehicles are on byperiodically classifying accelerometer samples collected by vehicles as theydrive in real time. Our key finding is that even adjacent patches of roadsurfaces contain characteristics that are sufficiently unique to differentiatebetween lanes, i.e., roads inherently exhibit differing bumps, cracks,potholes, and surface unevenness. Cars can capture this road surfaceinformation as they drive using inexpensive, easy-to-install accelerometersthat increasingly come fitted in cars and can be accessed via the CAN-bus. Wecollect an aggregate of 60 km driving data and synthesize more based on thisthat capture factors such as variable driving speed, vehicle suspensions, andaccelerometer noise. Our formulated LSTM-based deep learning model, LaNet,learns lane-specific sequences of road surface events (bumps, cracks etc.) andyields 100% lane classification accuracy with 200 meters of driving data,achieving over 90% with just 100 m (correspondingly to roughly one minute ofdriving). We design the LaNet model to be practical for use in real-time laneclassification and show with extensive experiments that LaNet yields highclassification accuracy even on smooth roads, on large multi-lane roads, and ondrives with frequent lane changes. Since different road surfaces have differentinherent characteristics or entropy, we excavate our neural network model anddiscover a mechanism to easily characterize the achievable classificationaccuracies in a road over various driving distances by training the model justonce. We present LaNet as a low-cost, easily deployable and highly accurate wayto achieve fine-grained lane identification.

LaNet: Real-time Lane Identification by Learning Road SurfaceCharacteristics from Accelerometer Data