An Efficient Approach Using Knowledge Distillation Methods to Stabilize Performance in a Lightweight Top-Down Posture Estimation Network

Abstract

Multi-person pose estimation has been gaining considerable interest due to its use in several real-world applications, such as activity recognition, motion capture, and augmented reality. Although the improvement of the accuracy and speed of multi-person pose estimation techniques has been recently studied, limitations still exist in balancing these two aspects. In this paper, a novel knowledge distilled lightweight top-down pose network (KDLPN) is proposed that balances computational complexity and accuracy. For the first time in multi-person pose estimation, a network that reduces computational complexity by applying a “Pelee” structure and shuffles pixels in the dense upsampling convolution layer to reduce the number of channels is presented. Furthermore, to prevent performance degradation because of the reduced computational complexity, knowledge distillation is applied to establish the pose estimation network as a teacher network. The method performance is evaluated on the MSCOCO dataset. Experimental results demonstrate that our KDLPN network significantly reduces 95% of the parameters required by state-of-the-art methods with minimal performance degradation. Moreover, our method is compared with other pose estimation methods to substantiate the importance of computational complexity reduction and its effectiveness.

Enhanced 3D Pose Estimation in Multi-Person, Multi-View Scenarios through Unsupervised Domain Adaptation with Dropout Discriminator

“…3D pose estimation from monocular inputs [14][15][16][17][18][19][20][21][22] presents an ill-posed problem, as multiple 3D predictions can correspond to the same 2D projection. Multi-view approaches have been developed to alleviate such projective ambiguity.…”

Multi-Person Pose Estimation Using Group-Based Convolutional Neural Network Model

“…Since the location and quantity of individuals in an image are uncertain, a multi-person pose estimate is more challenging than the single-person scenario. Typically both types of pose estimation can tackle the above issue using top-down, and bottom-up approach [6].…”