Search

When insufficient or incomplete multi-modality data are available in training, the corresponding classi-fication learning may lead to poor training performance or even failure. In order to tackle with this problem, the transfer learning algorithm called IMTL (incomplete modality transfer learning via latent low-rank constraint) is proposed in this paper. The proposed algorithm addresses the incomplete modality problem in two ways. Firstly, latent factors are introduced into a low-rank constrained subspace framework so as to mine missing modality infor-mation on the target domain. With the help of an auxiliary yet complete modality dataset, the proposed cross-modality and cross-dataset transfer learning strategy is used to help align data between modalities or datasets. Sec-ondly, a small amount of labeled target data is used to align the supervision information so as to maintain the internal structure of the target data during the transfer learning. Experimental results show that the proposed algorithm outperforms the previous transfer learning algorithms, and significantly improves the classification performance on the adopted incomplete target datasets.

Non-rigid point cloud registration holds significant importance for human body pose analysis in the fields of sports, medicine, gaming, etc. In this paper, we propose a non-rigid point cloud registration algorithm based on geodesic distance measurement, which can improve the accuracy of the registration for matching point pairs during non-rigid deformations. Firstly, a graph is constructed for two sets of point clouds using geodesic distance measurement considering that geodesic distance changes minimally during non-rigid deformation of the human body, which can preserve the point cloud matching information between corresponding points. Furthermore, a maximal clique search is employed to find combinations of matching pairs between point clouds. Finally, by driving the human body model parameters, sparse matching pairs are overlapped as much as possible to achieve non-rigid point cloud registration of the human body. The accuracy of the proposed algorithm is verified with FAUST and CAPE datasets. [ABSTRACT FROM AUTHOR]

For non-π-conjugated [SO 4 ] units, it is challenging to generate sufficient birefringence, owing to the high symmetry of the regular tetrahedron. Unlike the traditional trial-and-error approach, we propose a new paradigm for birefringence engineering to tune the optical properties based on [SO 4 ] units. Through the strategy of ligand substitution, we can predict its effect on the band gap and anisotropy. Theoretical evaluations reveal generalized results that the anisotropic electron distribution of new functional groups induced by the suitable ligand substitution contributes to the band gap and birefringence. To further validate the correctness of the paradigm, we experimentally synthesized and characterized nine novel compounds with selected functional modules. By the new paradigm of ligand substitution, they can reach up to 4-6 times the birefringence of the corresponding sulfate and maintain the wide bandgap. Through rational design, (CN 4 H 7 )SO 3 NH 2 exhibits about 35 times the birefringence of Li 2 SO 4 , which is a significant order of magnitude improvement and verifies the superiority of our proposed paradigm. This work provides a new paradigm for the modification to the non-π-conjugated group and will guide and accelerate the exploration of novel birefringent crystals in the short-wavelength region., Competing Interests: The authors declare no competing financial interest., (© 2024 The Authors. Published by American Chemical Society.)