Table 2 Quantitative comparison results using MPJPE (mm) with state-of-the-art methods on Human3.6M, using 2D ground truth key points as input.

From: Origin centric and part based pose decomposition for 3D human pose estimation

MPJPE

T

Dire.

Disc.

Eat

Greet

Phone

Photo

Pose

Pur.

Sit

SitD.

Smoke

Wait

WalkD.

Walk

WalkT.

Avg

VideoPose3D25

243

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

37.2

PoseFormer11

81

30.0

33.6

29.9

31.0

30.2

33.3

34.8

31.4

37.8

38.6

31.7

31.5

29.0

23.3

23.1

31.3

Stridedformer14

351

27.1

29.4

26.5

27.1

28.6

33.0

30.7

26.8

38.2

34.7

29.1

29.8

26.8

19.1

19.8

28.5

MixSTE13

243

21.6

22.0

20.4

21.0

20.8

24.3

24.7

21.9

26.9

24.9

21.2

21.5

20.8

14.7

15.7

21.6

HDFormer36

96

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

21.6

STCFormer35

243

20.8

21.8

20.0

20.6

23.4

25.0

23.6

19.3

27.8

26.1

21.6

20.6

19.5

14.3

15.1

21.3

MotionAGFormer16

243

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

17.3

MotionBERT18

243

16.7

19.9

17.1

16.5

17.4

18.8

19.3

20.5

24.0

22.1

18.6

16.8

16.7

10.8

11.5

17.8

Ours

243

14.2

15.4

15.0

14.0

15.5

17.7

16.3

16.2

21.0

20.0

16.2

13.7

8.9

14.2

9.7

15.2