IEEE Transactions on Visualization and Computer Graphics
Kangxue Yin^{1} Hui Huang^{2*} Edmond S. L. Ho^{3} Hao Wang^{2} Taku Komura^{4} Daniel CohenOr^{2,5} Hao Zhang^{1}
^{1}Simon Fraser University^{ } ^{2}Shenzhen University ^{3}Northumbria University ^{4}Edinburgh University ^{5}Tel Aviv University
Abstract

。。。。 

Figure 1. Closely interacting 3D wrestling poses automatically generated by MCMC sampling from a single seed pose (center).


Figure 2. Direct 2Dto3D lifting (second and third rows) vs. our approach (bottom) for the wrestling motion. Top row shows the annotated posepairs in video, as input to direct 2Dto3D lifting in second row [10] and third row [11]. Results in the bottom row are retrieved samples generated by our method, whose projections are close to the respective 2D posepairs in the top row. Quality and plausibility of the close interactions can be contrasted by focusing on the circled regions. 



Figure 3. Overview of our 3D posepair generation via liftingbysampling.
Figure 6. Ablation study of different terms in our probability density function that controls the MCMC sampling for Judo interactions. The 3D posepairs shown on the right, for each version of the PDF, represent the 200th, 400th, 600th, and 800th samples, respectively. To visually validate the results in (c), we also show examples of 2D annotations which contributed to the likelihood term in the bottom row (d).
Figure 7. With an unbiased proposal function, the sampling algorithm tends to produce more diverse but less plausible posepairs. The first row shows the 200th, 400th, 600th, and 800th samples generated from the same seed shown in Figure 6(a). The second row shows examples of 2D annotations that contribute to the likelihood term.

。。。 
Figure 10. Comparisons to enhanced lifting (b) and a naive sampling (c) around the 3D posepairs generated by enhanced lifting. For a clear assessment of the comparisons, we also show 2D annotations (a) and our results (d) that are the same as shown in Figure 2.




Figure 11. From a single seed, our sampling schema produces diverse posepairs of wrestling. Three closest 2D annotations are provided to the right side of each sampled 3D posepair to demonstrate its validity.
Bibtex
@article{Pose18,