人間と移動ロボットの共存

歩行者の移動には、環境歩行者自身の意図他者などが影響すると考えられます。これらの影響を考慮し、歩行者の移動を正確に予測することを目指しています。

周辺他者との関係を考慮した歩行者の移動予測

歩行者は周囲の他者からの影響を受けて移動しますが、この影響の度合いは歩行者間の関係によって異なります。このような周辺他者とのインタラクションを機械学習の方法を使って学習し、歩行者の予測を行う手法を提案しています。

Jiaxu Wu, Hanwool Woo, Yusuke Tamura, Alessandro Moro, Stefano Massaroli, Atsushi Yamashita, and Hajime Asama: Pedestrian Trajectory Prediction Using BiRNN Encoder-Decoder Framework, Advanced Robotics, Vol.33, No.18, pp.956-959, 2019. doi:10.1080/01691864.2019.1635910

abstract
Autonomous mobile robots navigating through human crowds are required to foresee the future trajectories of surrounding pedestrians and accordingly plan safe paths to avoid any possible collision. This paper presents a novel approach for pedestrian trajectory prediction. In particular, we developed a new method based on an encoder–decoder framework using bidirectional recurrent neural networks (BiRNN). The difficulty of incorporating social interactions into the model has been addressed thanks to the special structure of BiRNN enhanced by the attention mechanism, a proximity-independent model of the relative importance of each pedestrian. The main difference between our and the previous approaches is that BiRNN allows us to employs information on the future state of the pedestrians. We tested the performance of our method on several public datasets. The proposed model outperforms the current state-of-the-art approaches on most of these datasets. Furthermore, we analyze the resulting predicted trajectories and the learned attention scores to prove the advantages of BiRRNs on recognizing social interactions.

移動ロボットからの歩きスマホの検出

近年スマートフォンの普及に伴い、歩きスマホ(Smartphone Zombie)が社会的な問題となっています。スマートフォンに注意が集中することにより周辺への注意が疎かになり、他者と衝突する危険があります。このようなリスクの高い歩行者を移動ロボットに搭載したセンサ(LiDAR)によって検出する手法を提案しています。

Jiaxu Wu, Yusuke Tamura, Yusheng Wang, Hanwool Woo, Alessandro Moro, Atsushi Yamashita, and Hajime Asama: Smartphone Zombie Detection from LiDAR Point Cloud for Mobile Robot Safety, IEEE Robotics and Automation Letters, Vol.5, No.2, pp.2256-2263, 2020. doi:10.1109/LRA.2020.2970570

abstract
Awareness of surrounding and prediction of dangerous situations is essential for autonomous mobile robots, especially during navigation in a human-populated environment. To cope with safety issues, state-of-the-art works have focused on pedestrian detection, tracking, and trajectory prediction. However, only a few studies have been conducted on recognizing some specific types of dangerous behaviors exhibited by pedestrians. Here, we propose a tracking enhanced detection method to recognize people using their smartphones while walking, referred to as smartphone zombie. Features used for pedestrian detection usually involve the rotation variance problem, and in this paper, the drawback is handled by employing motion information from multi-object tracking. The proposed solution has been validated through experiments performed on a newly collected dataset. Results showed that our detector can learn a distinct pattern of the appearance of smartphone zombies. Thus, it can successfully detect them outperforming the existed detection method.


環境情報を考慮した歩行者の移動予測

例えば直線的な廊下を歩くとき、人間は廊下の向きに沿っておおよそ直線的に歩きます。このように、人間は環境の形状に影響を受けて移動します。また、例えばオフィス内で歩くときは、周辺にどのような部屋があるのか、エレベーターはどこか、トイレはどこか、など、環境の意味にも大きく影響を受けます。したがって、歩行者の移動を正確に予測するためには、このような環境の情報を考慮することが重要です。
我々は、環境中に設置したセンサ情報を利用することによって、その環境における歩行者の移動特性を蓄積し、移動予測に用いるという手法を提案しています。

田村雄介, 濱崎峻資, 山下淳, 淺間一: 環境に応じた人間の移動予測に基づく移動ロボットの人物回避, 日本機械学会論文集C編, Vol.79, No.799, pp.617-628, 2013. doi:10.1299/kikaic.79.617

abstract
In order for mobile robots to coexist with humans, both safety and efficiency should be satisfied. We propose a method to predict pedestrian movement for collision avoidance of mobile robots and pedestrians. In the proposed method, the pedestrian trajectories are measured and a database of human movement tendencies is generated. The database is applied to the prediction of future pedestrian movement. To decrease the initial time cost for database generation, environmental geometric configuration is considered in the form of virtual forces. To verify the usefulness of the proposed method, we generated the database based on five-hour observation and conducted three types of experiments based on the generated database. The first experiment showed the prediction performance of the proposed method and proved the method guaranteed the safety. The second experiment showed that the proposed method satisfied both the safety and efficiency through the comparative simulations. The third experiment showed the method could apply to the real mobile robot.

Shunsuke Hamasaki, Yusuke Tamura, Atsushi Yamashita and Hajime Asama: Prediction of Human’s Movement for Collision Avoidance of Mobile Robot, Proceedings of the 2011 IEEE International Conference on Robotics and Biomimetics, pp.1633-1638, Phuket, Thailand, December 2011. doi:10.1109/ROBIO.2011.6181523


歩行者の意図を考慮したモデル化

人間は、周囲に存在する他者や環境の状況から、周辺他者を回避するか、追従するか、など意図を切り替えながら歩行していると考えられます。我々は、Social Force Model1 を拡張し、このような歩行意図の切り替えに応じたサブゴールを生成しながら動作生成を行うモデルを提案しています。


  1. 歩行者を質点とみなし、環境、他者、目的地などから仮想的な力が働くことによって移動するというモデル (Helbing & Molnár, 1995)。