arXiv Science⌕ Search

arXiv subjects

Keyu Chen

Publications and source records attributed to Keyu Chen.

112 records · Page 7Linked to original sources

Unknown Identity Rejection Loss: Utilizing Unlabeled Data for Face Recognition

Face recognition has advanced considerably with the availability of large-scale labeled datasets. However, how to further improve the performance with the easily accessible unlabeled dataset remains a challenge. In this paper, we propose the novel Unknown Identity Rejection (UIR) loss to utilize the unlabeled data. We categorize identities in unconstrained environment into the known set and the unknown set. The former corresponds to the identities that appear in the labeled training dataset while the latter is its complementary set. Besides training the model to accurately classify the known identities, we also force the model to reject unknown identities provided by the unlabeled dataset via our proposed UIR loss. In order to 'reject' faces of unknown identities, centers of the known identities are forced to keep enough margin from centers of unknown identities which are assumed to be approximated by the features of their samples. By this means, the discriminativeness of the face representations can be enhanced. Experimental results demonstrate that our approach can provide obvious performance improvement by utilizing the unlabeled data.

cs.CV↗

Pressure-Induced Large Volume Collapse, Plane-to-Chain, Insulator to Metal Transition in CaMn$_2$Bi$_2$

In-situ high pressure single crystal X-ray diffraction study reveals that the quantum material CaMn$_2$Bi$_2$ undergoes a unique plane to chain structural transition between 2 and 3 GPa, accompanied by a large volume collapse. CaMn2Bi2 displays a new structure type above 2.3 GPa, with the puckered Mn honeycomb lattice of the trigonal ambient-pressure structure converting to one-dimensional (1D) zigzag chains in the high-pressure monoclinic structure. Single crystal measurements reveal that the pressure-induced structural transformation is accompanied by a dramatic two order of magnitude drop of resistivity; although the ambient pressure phase displays semiconducting behavior at low temperatures, metallic temperature dependent resistivity is observed for the high pressure phase, as, surprisingly, are two resistivity anomalies with opposite pressure dependences. Based on the electronic structure calculations, we hypothesized that the newly emerged electronic state under high pressure is associated with a Fermi surface instability of the quasi-1D Mn chains, while we infer that the other is a magnetic transition. Assessment of the total energies for hypothetical magnetic structures for high pressure CaMn$_2$Bi$_2$ indicates that ferrimagnetism is thermodynamically favored.

cond-mat.mtrl-sci↗

Disentangled Representation Learning for 3D Face Shape

In this paper, we present a novel strategy to design disentangled 3D face shape representation. Specifically, a given 3D face shape is decomposed into identity part and expression part, which are both encoded and decoded in a nonlinear way. To solve this problem, we propose an attribute decomposition framework for 3D face mesh. To better represent face shapes which are usually nonlinear deformed between each other, the face shapes are represented by a vertex based deformation representation rather than Euclidean coordinates. The experimental results demonstrate that our method has better performance than existing methods on decomposing the identity and expression parts. Moreover, more natural expression transfer results can be achieved with our method than existing methods.

cs.CV↗

Image Matters: Visually modeling user behaviors using Advanced Model Server

In Taobao, the largest e-commerce platform in China, billions of items are provided and typically displayed with their images. For better user experience and business effectiveness, Click Through Rate (CTR) prediction in online advertising system exploits abundant user historical behaviors to identify whether a user is interested in a candidate ad. Enhancing behavior representations with user behavior images will help understand user's visual preference and improve the accuracy of CTR prediction greatly. So we propose to model user preference jointly with user behavior ID features and behavior images. However, training with user behavior images brings tens to hundreds of images in one sample, giving rise to a great challenge in both communication and computation. To handle these challenges, we propose a novel and efficient distributed machine learning paradigm called Advanced Model Server (AMS). With the well known Parameter Server (PS) framework, each server node handles a separate part of parameters and updates them independently. AMS goes beyond this and is designed to be capable of learning a unified image descriptor model shared by all server nodes which embeds large images into low dimensional high level features before transmitting images to worker nodes. AMS thus dramatically reduces the communication load and enables the arduous joint training process. Based on AMS, the methods of effectively combining the images and ID features are carefully studied, and then we propose a Deep Image CTR Model. Our approach is shown to achieve significant improvements in both online and offline evaluations, and has been deployed in Taobao display advertising system serving the main traffic.

cs.CV↗