OccRobNet : Occlusion Robust Network for Accurate 3D Interacting Hand-Object Pose Estimation
This work addresses occlusion challenges in 3D hand pose estimation for applications like robotics and VR, but it is incremental as it builds on existing methods with specific improvements.
The paper tackles the problem of occlusion in 3D hand pose estimation during hand-object interactions by proposing OccRobNet, which uses self-attention transformers to identify joints in occluded regions, achieving state-of-the-art results on InterHand2.6M, HO3D, and H2O3D datasets.
Occlusion is one of the challenging issues when estimating 3D hand pose. This problem becomes more prominent when hand interacts with an object or two hands are involved. In the past works, much attention has not been given to these occluded regions. But these regions contain important and beneficial information that is vital for 3D hand pose estimation. Thus, in this paper, we propose an occlusion robust and accurate method for the estimation of 3D hand-object pose from the input RGB image. Our method includes first localising the hand joints using a CNN based model and then refining them by extracting contextual information. The self attention transformer then identifies the specific joints along with the hand identity. This helps the model to identify the hand belongingness of a particular joint which helps to detect the joint even in the occluded region. Further, these joints with hand identity are then used to estimate the pose using cross attention mechanism. Thus, by identifying the joints in the occluded region, the obtained network becomes robust to occlusion. Hence, this network achieves state-of-the-art results when evaluated on the InterHand2.6M, HO3D and H$_2$O3D datasets.