Jian Xu (徐健)

✉ jian.xu@ia.ac.cn  

I am currently an Associate Professor at Institute of Automation Chinese Academy of Sciences (CASIA).

Before joining CASIA, I have 3 years of experience in AI corporations HUAWEI and XREAL.

I obtained my Ph.D. in Pattern Recognition and Intelligent Systems from Institute of Automation, Chinese Academy of Science in 2020, under the supervision of Prof. Chunheng Wang. Previously I received my B.S. in Control Science and Engineering from Shandong University in 2015.

My research interests lie in Large Multimodal Models, AI4Science, Pose Estimation and Image Retrieval.

A full-body human motion dataset that captures text-guided desktop object rearrangement through MoCap and AR glasses; & A pipeline for generating avatar's motion of object rearrangement driven by text instruction.

A single-view hand-held object reconstruction method that exploits the categorical shape prior to reconstruct the shape of intra-class objects; & A new synthetic dataset, COMIC, that contains the category-level collection of objects with diverse shape, materials, interacting poses, and viewing directions.

We propose a multi-view hand mesh recovery (HMR) method with Transformer. It leverages the "power of points", including Basis Points Set, point's positional encoding and point-Transformer, to unify and merge information from sparsely arranged cameras.

We propose to directly predict object-level depth from a monocular RGB image by deforming the category-level shape prior into object-level depth and the canonical NOCS representation.

We propose a simple but effective semantic-based aggregation (SBA) method.

We propose the iterative manifold embedding (IME) layer, of which the weights are learned offline by an unsupervised strategy, to explore the intrinsic manifolds by incomplete data.

We propose a simple but effective semantic part-based weighting aggregation (PWA) for image retrieval.

The website template was adapted from Xingyu Chen.