IMoS: Intent‐Driven Full‐Body Motion Synthesis for Human‐Object Interactions

Ghosh, Anindita; Dabral, Rishabh; Golyanik, Vladislav; Theobalt, Christian; Slusallek, Philipp

doi:10.1111/cgf.14739

Cited by 25 publications

(1 citation statement)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Traditional studies have focused on techniques for detecting 3D objects [GD07,GSEH11] and predicting affordances based on human poses [DFL*12,GGVG11, FDG*12]. In more recent works, there has been an emphasis on generating realistic static poses within the context of a 3D scene [LLK*19, ZBT21, ZZM*20, HGT*21, ZWZ*22, HWL*23, GDG*23], leveraging newly collected datasets on human interactions [HCTB19, GMSPM21, BXP*22,TGBT20]. Compared to the studies focused on creating static poses, our work tackles a task of generating coherent motions that align with the scene.…”

Section: Related Workmentioning

confidence: 99%

DAFNet: Generating Diverse Actions for Furniture Interaction by Learning Conditional Pose Distribution

Jin,

Lee

2023

Computer Graphics Forum

View full text Add to dashboard Cite

We present DAFNet, a novel data‐driven framework capable of generating various actions for indoor environment interactions. By taking desired root and upper‐body poses as control inputs, DAFNet generates whole‐body poses suitable for furniture of various shapes and combinations. To enable the generation of diverse actions, we introduce an action predictor that automatically infers the probabilities of individual action types based on the control input and environment. The action predictor is learned in an unsupervised manner by training Gaussian Mixture Variational Autoencoder (GMVAE). Additionally, we propose a two‐part normalizing flow‐based pose generator that sequentially generates upper and lower body poses. This two‐part model improves motion quality and the accuracy of satisfying conditions over a single model generating the whole body. Our experiments show that DAFNet can create continuous character motion for indoor scene scenarios, and both qualitative and quantitative evaluations demonstrate the effectiveness of our framework.We propose DAFNet, a novel data‐driven framework that can generate various actions for indoor environment interactions. Given the desired root and upper‐body pose as control inputs, DAFNet generates whole‐body poses for a character appropriate for furniture of various shapes and combinations.image

show abstract