An Improved Tiered Head Pose Estimation Network with Self-Adjust Loss Function

Abstract

As an important task in computer vision, head pose estimation has been widely applied in both academia and industry. However, there remains two challenges in the field of head pose estimation: (1) even given the same task (e.g., tiredness detection), the existing algorithms usually consider the estimation of the three angles (i.e., roll, yaw, and pitch) as separate facets, which disregard their interplay as well as differences and thus share the same parameters for all layers; and (2) the discontinuity in angle estimation definitely reduces the accuracy. To solve these two problems, a THESL-Net (tiered head pose estimation with self-adjust loss network) model is proposed in this study. Specifically, first, an idea of stepped estimation using distinct network layers is proposed, gaining a greater freedom during angle estimation. Furthermore, the reasons for the discontinuity in angle estimation are revealed, including not only labeling the dataset with quaternions or Euler angles, but also the loss function that simply adds the classification and regression losses. Subsequently, a self-adjustment constraint on the loss function is applied, making the angle estimation more consistent. Finally, to examine the influence of different angle ranges on the proposed model, experiments are conducted on three popular public benchmark datasets, BIWI, AFLW2000, and UPNA, demonstrating that the proposed model outperforms the state-of-the-art approaches.

Attention-Guided Huber Loss for Head Pose Estimation Based on Improved Capsule Network

“…FDN [38] uses a feature decoupling model to explicitly identify the discriminative features for each angle by adaptively recalibrating the channel responses of each pose angle and suppressing less useful features. Zhu [39] proposed a hierarchical estimation method based on distinct network layers, gaining greater degrees of freedom in the angle estimation process. The LwPosr [40] uses a mixture of depth-separable convolution and Transformer [41] encoder layers with a dual-stream heterogeneous structure to extract features to provide fine-grained regression for predicting head pose.…”

“…The visualization analysis results indicate that our proposed method is closer to the ground truth labels in cases of significant angle deviation and occlusions compared to FSA-Net. In addition, to strengthen the persuasive power of our experiments, we refer to the visualization results of Zhu [39]. Based on their work, we compare our proposed method with other methods.…”

“…The results in Figure 12 show that our method achieves the best performance. In addition, to strengthen the persuasive power of our experiments, we refer to the visualization results of Zhu [39]. Based on their work, we compare our proposed method with other methods.…”