THÖR-Magni: Comparative Analysis of Deep Learning Models for Role-Conditioned Human Motion Prediction

Almeida, Tiago,1996-Örebro universitet,Institutionen för naturvetenskap och teknik (author)

THÖR-Magni : Comparative Analysis of Deep Learning Models for Role-Conditioned Human Motion Prediction

Autonomous systems, that need to operate in human environments and interact with the users, rely on understanding and anticipating human activity and motion. Among the many factors which influence human motion, semantic attributes, such as the roles and ongoing activities of the detected people, provide a powerful cue on their future motion, actions, and intentions. In this work we adapt several popular deep learning models for trajectory prediction with labels corresponding to the roles of the people. To this end we use the novel THOR-Magni dataset, which captures human activity in industrial settings and includes the relevant semantic labels for people who navigate complex environments, interact with objects and robots, work alone and in groups. In qualitative and quantitative experiments we show that the role-conditioned LSTM, Transformer, GAN and VAE methods can effectively incorporate the semantic categories, better capture the underlying input distribution and therefore produce more accurate motion predictions in terms of Top-K ADE/FDE and log-likelihood metrics.

Rudenko, Andrey,1991-Robert Bosch GmbH, Corporate Research, Stuttgart, Germany (author)
Schreiter, Tim,1997-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)tmsr (author)
Zhu, Yufei,1994-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)yzu (author)
Gutiérrez Maestro, Eduardo,1994-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)ego (author)
Morillo-Mendez, Lucas,1991-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)lsmo (author)
Kucner, Tomasz P.Mobile Robotics Group, Department of Electrical Engineering and Automation, Aalto University, Finland; FCAI, Finnish Center for Artificial Intelligence, Finland (author)
Martinez Mozos, Oscar,1974-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)oms (author)
Magnusson, Martin,Docent,1977-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)mnmn (author)
Palmieri, LuigiRobert Bosch GmbH, Corporate Research, Stuttgart, Germany (author)
Arras, Kai O.Robert Bosch GmbH, Corporate Research, Stuttgart, Germany (author)
Lilienthal, Achim,1970-Örebro universitet,Institutionen för naturvetenskap och teknik(Swepub:oru)amll (author)
Örebro universitetInstitutionen för naturvetenskap och teknik (creator_code:org_t)

In:2023 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW): IEEE, s. 2192-220197983503074509798350307443

By the author/editor: Almeida, Tiago, ...; Rudenko, Andrey, ...; Schreiter, Tim, ...; Zhu, Yufei, 1994 ...; Gutiérrez Maestr ...; Morillo-Mendez, ...; show more...; Kucner, Tomasz P ...; Martinez Mozos, ...; Magnusson, Marti ...; Palmieri, Luigi; Arras, Kai O.; Lilienthal, Achi ...; show less...

About the subject

NATURAL SCIENCES: NATURAL SCIENCES; and Computer and Inf ...; and Computer Vision ...

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.