Research Publications People Teaching Position Openings

Research Grants:

Pujiang Talent Program(浦江人才计划), PI, 2015-2017

Shanghai Outstanding academic leaders (上海市优秀学术带头人),co-PI(第二责任人和重点培养对象)

National Natural Science Foundation of China (国家自然科学基金青年基金), PI, 2016-2018



ACM Shanghai Young Research Scientist Award, 2015

Microsoft Research Fellowship, 2010



tutorial @ VCIP'2015, "Regularities of Visual Data and Their Applications"

Tutorial @ACCV'2014, "Advanced Sparse Representation for image and video analaysis"


Research Directions:

1) Object recognition  We focus on proposing more robust and effective image representations for image representation.

Laplacian sparse coding: We leverage the dependency among the local features for better sparse representation. [link]
Kernel sparse representation: Nonlinearity structure of the local features are considered sparse representation in Reproducing Kernel Hilbert Space [link]
Multi-Layer group sparse coding: Image tagging and classifications are jointly conducted in multi-layer group sparse coding framework.[link]
Category-Specific and Shared Dictionary Learning: A novel dictionary learning algorithm is proposed to exploit the commonality and Specificity among different object categories for better object representation. [link]

2)  Face recognition
Single Sample per Person Face Recognition: We proposed Regularized patch-based Representation for Single Sample Per Person Face Recognition. Such algorithm inherits the advantages of both holistic face representation and patch-based Face Representation.[link]
Image Set based Face Recognition: A patch-based sparse representation is posed for alignment-free image set classification. [link]

3)Deep leaning algorithm

DEFEATnet: We use sparse coding as the basic building block to build the deep neural network for better image representation. [link]
PCAnet: We use the simplest PCA as the components to build deep structure. Such framework achieves the best performance on the FERET dataset. [link]