ترغب بنشر مسار تعليمي؟ اضغط هنا

We are More than Our Joints: Predicting how 3D Bodies Move

116   0   0.0 ( 0 )
 نشر من قبل Yan Zhang
 تاريخ النشر 2020
  مجال البحث الهندسة المعلوماتية
والبحث باللغة English




اسأل ChatGPT حول البحث

A key step towards understanding human behavior is the prediction of 3D human motion. Successful solutions have many applications in human tracking, HCI, and graphics. Most previous work focuses on predicting a time series of future 3D joint locations given a sequence 3D joints from the past. This Euclidean formulation generally works better than predicting pose in terms of joint rotations. Body joint locations, however, do not fully constrain 3D human pose, leaving degrees of freedom undefined, making it hard to animate a realistic human from only the joints. Note that the 3D joints can be viewed as a sparse point cloud. Thus the problem of human motion prediction can be seen as point cloud prediction. With this observation, we instead predict a sparse set of locations on the body surface that correspond to motion capture markers. Given such markers, we fit a parametric body model to recover the 3D shape and pose of the person. These sparse surface markers also carry detailed information about human movement that is not present in the joints, increasing the naturalness of the predicted motions. Using the AMASS dataset, we train MOJO, which is a novel variational autoencoder that generates motions from latent frequencies. MOJO preserves the full temporal resolution of the input motion, and sampling from the latent frequencies explicitly introduces high-frequency components into the generated motion. We note that motion prediction methods accumulate errors over time, resulting in joints or markers that diverge from true human bodies. To address this, we fit SMPL-X to the predictions at each time step, projecting the solution back onto the space of valid bodies. These valid markers are then propagated in time. Experiments show that our method produces state-of-the-art results and realistic 3D body animations. The code for research purposes is at https://yz-cnsdqz.github.io/MOJO/MOJO.html



قيم البحث

اقرأ أيضاً

We report the discovery of two new giant radio galaxies (GRGs) using the MeerKAT International GHz Tiered Extragalactic Exploration (MIGHTEE) survey. Both GRGs were found within a 1 deg^2 region inside the COSMOS field. They have redshifts of z=0.165 6 and z=0.3363 and physical sizes of 2.4Mpc and 2.0Mpc, respectively. Only the cores of these GRGs were clearly visible in previous high resolution VLA observations, since the diffuse emission of the lobes was resolved out. However, the excellent sensitivity and uv coverage of the new MeerKAT telescope allowed this diffuse emission to be detected. The GRGs occupy a unpopulated region of radio power - size parameter space. Based on a recent estimate of the GRG number density, the probability of finding two or more GRGs with such large sizes at z<0.4 in a ~1deg^2 field is only 2.7x10^-6, assuming Poisson statistics. This supports the hypothesis that the prevalence of GRGs has been significantly underestimated in the past due to limited sensitivity to low surface brightness emission. The two GRGs presented here may be the first of a new population to be revealed through surveys like MIGHTEE which provide exquisite sensitivity to diffuse, extended emission.
122 - Sophia U. David 2020
When presented with information of any type, from music to language to mathematics, the human mind subconsciously arranges it into a network. A network puts pieces of information like musical notes, syllables or mathematical concepts into context by linking them together. These networks help our minds organize information and anticipate what is coming. Here we present two questions about network building. 1) Can humans more easily learn some types of networks than others? 2) Do humans find some links between ideas more surprising than others? The answer to both questions is Yes, and we explain why. The findings provide much-needed insight into the ways that humans learn about the networked world around them. Moreover, the study paves the way for future efforts seeking to optimize how information is presented to accelerate human learning.
129 - Michael S. Turner 2021
The $rmLambda$CDM cosmological model is remarkable: with just 6 parameters it describes the evolution of the Universe from a very early time when all structures were quantum fluctuations on subatomic scales to the present, and it is consistent with a wealth of high-precision data, both laboratory measurements and astronomical observations. However, the foundation of $rmLambda$CDM involves physics beyond the standard model of particle physics: particle dark matter, dark energy and cosmic inflation. Until this `new physics is clarified, $rmLambda$CDM is at best incomplete and at worst a phenomenological construct that accommodates the data. I discuss the path forward, which involves both discovery and disruption, some grand challenges and finally the limits of scientific cosmology.
Modeling how human moves in the space is useful for policy-making in transportation, public safety, and public health. Human movements can be viewed as a dynamic process that human transits between states (eg, locations) over time. In the human world where intelligent agents like humans or vehicles with human drivers play an important role, the states of agents mostly describe human activities, and the state transition is influenced by both the human decisions and physical constraints from the real-world system (eg, agents need to spend time to move over a certain distance). Therefore, the modeling of state transition should include the modeling of the agents decision process and the physical system dynamics. In this paper, we propose ours to model state transition in human movement from a novel perspective, by learning the decision model and integrating the system dynamics. ours learns the human movement with Generative Adversarial Imitation Learning and integrates the stochastic constraints from system dynamics in the learning process. To the best of our knowledge, we are the first to learn to model the state transition of moving agents with system dynamics. In extensive experiments on real-world datasets, we demonstrate that the proposed method can generate trajectories similar to real-world ones, and outperform the state-of-the-art methods in predicting the next location and generating long-term future trajectories.
In cite{CGH15} we introduced TiRS graphs and TiRS frames to create a new natural setting for duals of canonical extensions of lattices. In this continuation of cite{CGH15} we answer Problem 2 from there by characterising the perfect lattices that are dual to TiRS frames (and hence TiRS graphs). We introduce a new subclass of perfect lattices called PTi lattices and show that the canonical extensions of lattices are PTi lattices, and so are `more than just perfect lattices. We introduce morphisms of TiRS structures and put our correspondence between TiRS graphs and TiRS frames from cite{CGH15} into a full categorical framework. We illustrate our correspondences between classes of perfects lattices and classes of TiRS graphs by examples.
التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا