Eight-coordinate mono- and dinuclear Dy(III) complexes containing a rigid equatorial plane and an anisobidentate carboxylate ligand in the axial position: synthesis, structure and magnetism.

A rigid pentadentate chelating ligand (HL) has been utilized to synthesize a series of octacoordinate mononuclear complexes, [Dy(L)(PhPO)(OOCR)] (where R = CH (1), C(CH) (2), CF (3)) and a dinuclear complex, [Dy(L)(PhPO){(OOC)CH}] (4) based on the highly anisotropic Dy(III) ion. All the complexes were structurally characterized by single-crystal X-ray diffraction studies. The complexes were formed by the coordination action of the dianionic pentadentate ligand [L], one phosphine oxide, and carboxylate ligands.

PoseScript: Linking 3D Human Poses and Natural Language.

Natural language plays a critical role in many computer vision applications, such as image captioning, visual question answering, and cross-modal retrieval, to provide fine-grained semantic information. Unfortunately, while human pose is key to human understanding, current 3D human pose datasets lack detailed language descriptions. To address this issue, we have introduced the PoseScript dataset.

Incorporation of Photo- and Thermoresponsive -Salicylidene Aniline Derivatives into Cobalt and Zinc Layered Hydroxides.

In search of new multifunctional hybrid materials and in order to investigate the influence of chemical modification on the possible synergy between properties, the carboxylate and sulfonate derivatives of photo- and thermochromic N-salicylidene aniline were successfully inserted into Co(II)- and Zn(II)-based layered simple hydroxides, resulting in four novel hybrids: Co--Sali-COO, Co--Sali-SO, Zn--Sali-COO, and Zn--Sali-SO. All synthesized hybrids adopt a double organic layered configuration, which prevents the photoisomerization ability of -Sali-R molecules in the hybrids. However, the Zn hybrids exhibit fluorescence upon exposure to UV light due to the excited-state intramolecular proton transfer (ESIPT) mechanism.

PoseBERT: A Generic Transformer Module for Temporal 3D Human Modeling.

Training state-of-the-art models for human pose estimation in videos requires datasets with annotations that are really hard and expensive to obtain. Although transformers have been recently utilized for body pose sequence modeling, related methods rely on pseudo-ground truth to augment the currently limited training data available for learning such models. In this paper, we introduce PoseBERT, a transformer module that is fully trained on 3D Motion Capture (MoCap) data via masked modeling.