Hao Xu

Hao Xu, Haipeng Li, Yinqiao Wang et al.

Reconstructing 3D hand mesh robustly from a single image is very challenging, due to the lack of diversity in existing real-world datasets. While data synthesis helps relieve the issue, the syn-to-real gap still hinders its usage. In this work, we present HandBooster, a new approach to uplift the data diversity and boost the 3D hand-mesh reconstruction performance by training a conditional generative space on hand-object interactions and purposely sampling the space to synthesize effective data samples. First, we construct versatile content-aware conditions to guide a diffusion model to produce realistic images with diverse hand appearances, poses, views, and backgrounds; favorably, accurate 3D annotations are obtained for free. Then, we design a novel condition creator based on our similarity-aware distribution sampling strategies to deliberately find novel and realistic interaction poses that are distinctive from the training set. Equipped with our method, several baselines can be significantly improved beyond the SOTA on the HO3D and DexYCB benchmarks. Our code will be released on https://github.com/hxwork/HandBooster_Pytorch.

Shangyuan Li, Li Feng, Deqiang Gan et al.

Load model identification using small disturbance data is studied. It is proved that the individual load to be identified and the rest of the system forms a closed-loop system. Then, the impacts of disturbances entering the feedforward channel (internal disturbance) and feedback channel (external disturbance) on relationship between load inputs and outputs are examined analytically. It is found out that relationship between load inputs and outputs is not determined by load itself (feedforward transfer function) only, but also related with equivalent network matrix (feedback transfer function). Thus, load identification is closed loop identification essentially and the impact of closed loop identification cannot be neglected when using small disturbance data to identify load parameters. Closed loop load model identification can be solved by prediction error method (PEM). Implementation of PEM based on a Kalman filtering formulation is detailed. Identification results using simulated data demonstrates the correctness and significance of theoretical analysis.