Lane changing behavior recognition based on Artificial Neural Network-based State Machine approach
DOI: 10.1109/itsc48978.2021.9564919
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the challenge of accurately recognizing and predicting driver lane-changing behaviors for Advanced Driving Assistance Systems (ADAS). Motivated by the fact that human error causes the majority of traffic accidents, the authors aim to develop a model that is both highly accurate and interpretable, overcoming the "black box" nature of conventional Machine Learning approaches. The study proposes an Artificial Neural Network (ANN)-based State Machine approach to estimate three specific driving states: lane keeping, lane change to the right, and lane change to the left. The methodology combines a state machine topology, which defines discrete driving states and transitions, with ANNs that estimate the conditions for these transitions. Two concepts are evaluated: Approach I uses a single common ANN for all states, while Approach II employs three distinct ANNs, each corresponding to a specific driving state. The model utilizes operational inputs (steering angle, pedal positions, indicators) and environmental inputs (time-to-collision with surrounding vehicles). To optimize the ANN weights and biases, the authors employ the Non-dominated Sorting Genetic Algorithm II (NSGA-II). Experimental data were collected from three participants using a driving simulator, with separate datasets for training and testing to assess model generalizability. The results demonstrate that the proposed ANN-based State Machine approaches significantly outperform conventional ANN models. Specifically, Approach II achieved the highest performance, with mean accuracy rates exceeding 92% across all test datasets. The study reports improvements of up to 46% in detection rates and 34% in false alarm rates compared to conventional ANNs. Detailed metrics show high accuracy for right lane changes (up to 98.59%) and lane keeping, with generally low false alarm rates. The model successfully maintained high performance when tested on data from drivers not included in the training set, proving its generalizability. The significance of this work lies in providing a trainable and interpretable framework for driving behavior recognition. By integrating ANNs with a state machine, the model offers the predictive power of neural networks while maintaining the structural clarity of state-based logic. This approach enables ADAS to better understand individual driving habits and predict maneuvers with high accuracy and low false alarms, contributing to improved road safety and more effective driver assistance systems.
Key finding
The Artificial Neural Network-based state machine approach achieves significantly higher detection rates and lower false alarm rates compared to conventional Artificial Neural Networks for recognizing lane changing behaviors.
Methodology
simulator
Sample size: 3
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 7 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Theoretical Contribution: computational model