2022 Fiscal Year Final Research Report
Behavior Prediction of Traffic Participants based on Micro and Macro features for Urban Automated Driving
Project/Area Number |
20K04397
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Review Section |
Basic Section 20020:Robotics and intelligent system-related
|
Research Institution | Kanazawa University |
Principal Investigator |
Yoneda Keisuke 金沢大学, 新学術創成研究機構, 准教授 (80643957)
|
Co-Investigator(Kenkyū-buntansha) |
菅沼 直樹 金沢大学, 新学術創成研究機構, 教授 (50361978)
倉元 昭季 東京都立大学, システムデザイン研究科, 助教 (90826851)
|
Project Period (FY) |
2020-04-01 – 2023-03-31
|
Keywords | 自動運転自動車 / 行動予測 / 移動ロボット / 深層学習 / 画像処理 |
Outline of Final Research Achievements |
In the research on urban automated driving, surrounding object recognition utilizing onboard sensors’ information is one of the important technologies. Automated vehicles have to generate safe driving behaviors taking into account the movements and intentions of surrounding traffic participants such as vehicles, pedestrians, and cyclists. This research project investigates a behavior prediction method considering both apparent and latent behaviors of objects. We developed behavior prediction methods that can integrate multiple behavior models. In addition, different behavior prediction models are designed including a behavior model that predicts movements around the environment from a bird's-eye view, a behavior model that predicts the interaction between objects based on objects’ motion and a digital map, and a pedestrian posture estimation model based on camera images.
|
Free Research Field |
情報科学
|
Academic Significance and Societal Importance of the Research Achievements |
本研究では自動運転自動車が遭遇する交差点走行に注目し周辺との相互作用をモデル化したミクロな行動予測技術及び周辺の交通参加者の潜在的な動きを俯瞰的に予測するマクロな行動予測技術を開発し,これらを統合した予測精度改善の実現を目的と設定した.個別の物体を中心としたエージェントベースの行動予測及び俯瞰的な予測の双方の視点から横断的に統合し,物体の潜在的な動きの変化を考慮した滑らかな物体予測技術の実現を目指している.本研究の達成により,一般ドライバが感覚的に行う予測技術を自動運転の機能として実現することに貢献可能である.交差点走行の状況予測が強化され安全な走行環境の確保に期待したい.
|