MoPO: Incorporating Motion Prior for Occluded Human Mesh Recovery

Dimensions overall score 9.0

GitHub Code Pulse

No public code linked for this paper yet.

• MoPO incorporates a motion de-occlusion module that uses a spatial-temporal occlusion detector to detect joint visibility and a lightweight motion predictor to complete occluded body parts based on history poses.method 95%
  Evidencepartial

  The motion de-occlusion module, where we propose a spatial-temporal occlusion detector to detect joint visibility, and then we propose a lightweight motion predictor to complete the occluded body parts by predicting the most plausible joint positions based on history poses.

  Implicationpartial

  Directly stated in the abstract and described as a key component of the method.

  Verificationpartial

  partial

• MoPO includes a motion-aware fusion and refinement module that fuses the completed joint sequence with image features to estimate human shape and initial human pose, and uses the completed joint sequence to refine the final human pose through inverse kinematics.method 95%
  Evidencepartial

  The motion-aware fusion and refinement module, which fuses the completed joint sequence with image features to estimate human shape and initial human pose. Moreover, the completed joint sequence is further used to refine the final human pose through inverse kinematics, which provides the occlusion-free motion prior for regressing human poses.

  Implicationpartial

  Directly stated in the abstract as a core component of the method.

  Verificationpartial

  partial

• MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks.result 90%
  Evidencepartial

  Extensive experiments demonstrate that MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks, significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery.

  Implicationpartial

  Directly stated in the abstract, but specific benchmark names and numeric results are not provided in the excerpt.

  Verificationpartial

  partial

• MoPO significantly enhances the accuracy and temporal consistency of occluded human mesh recovery.result 90%
  Evidencepartial

  significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery.

  Implicationpartial

  Directly stated in the abstract as a key result.

  Verificationpartial

  partial

• Pose sequence inherently contains reliable motion prior for estimating occluded body parts compared to occluded image features.technical 85%
  Evidencepartial

  we discover that compared to occluded image features, pose sequence inherently contains reliable motion prior for estimating occluded body parts.

  Implicationpartial

  Stated as a discovery in the abstract, but the evidence is qualitative rather than quantitative.

  Verificationpartial

  partial

• Previous human mesh recovery methods exhibit limited robustness to occlusions and produce inaccurate poses and severe motion jitter.limitation 80%
  Evidencepartial

  they still exhibit limited robustness to occlusions and often produce inaccurate poses and severe motion jitter due to the insufficient spatial features for occluded body parts.

  Implicationpartial

  Stated as a limitation of prior work in the abstract, but no specific prior methods or quantitative comparisons are given in the excerpt.

  Verificationpartial

  partial

• MoPO introduces a motion de-occlusion module that uses a spatial-temporal occlusion detector to detect joint visibility and a lightweight motion predictor to complete occluded body parts based on history poses.method 95%
  Evidencepartial

  The motion de-occlusion module, where we propose a spatial-temporal occlusion detector to detect joint visibility, and then we propose a lightweight motion predictor to complete the occluded body parts by predicting the most plausible joint positions based on history poses.

  Implicationpartial

  Directly stated in the abstract with clear description of components.

  Verificationpartial

  partial

• MoPO includes a motion-aware fusion and refinement module that fuses completed joint sequences with image features to estimate human shape and initial pose, and uses inverse kinematics to refine the final pose.method 95%
  Evidencepartial

  The motion-aware fusion and refinement module, which fuses the completed joint sequence with image features to estimate human shape and initial human pose. Moreover, the completed joint sequence is further used to refine the final human pose through inverse kinematics, which provides the occlusion-free motion prior for regressing human poses.

  Implicationpartial

  Directly stated in the abstract.

  Verificationpartial

  partial

• MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks for human mesh recovery.result 90%
  Evidencepartial

  Extensive experiments demonstrate that MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks, significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery.

  Implicationpartial

  Explicitly stated in the abstract, but specific benchmark numbers are not provided in the excerpt.

  Verificationpartial

  partial

• MoPO significantly enhances temporal consistency in occluded human mesh recovery.result 85%
  Evidencepartial

  significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery.

  Implicationpartial

  Directly stated in the abstract, but without specific quantitative metrics in the excerpt.

  Verificationpartial

  partial

• Pose sequences inherently contain reliable motion prior for estimating occluded body parts compared to occluded image features.technical 80%
  Evidencepartial

  we discover that compared to occluded image features, pose sequence inherently contains reliable motion prior for estimating occluded body parts.

  Implicationpartial

  Stated as a discovery in the abstract, but not backed by specific evidence in the excerpt.

  Verificationpartial

  partial

• Previous human mesh recovery methods exhibit limited robustness to occlusions and produce inaccurate poses and severe motion jitter.limitation 70%
  Evidencepartial

  they still exhibit limited robustness to occlusions and often produce inaccurate poses and severe motion jitter due to the insufficient spatial features for occluded body parts.

  Implicationpartial

  Stated as a limitation of prior work, but no specific references or data are provided in the excerpt.

  Verificationpartial

  partial

Startup potential card