Tian Wang

Dr. Tian Wang

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 November 2023
Verified
Conference Paper
Closed access
TPTaiying Peng
Taiying Peng
TWTian Wang
Tian Wang
DLDeyuan Liu
Deyuan Liu
JWJian Wang
Jian Wang
YFYao Fu
Yao Fu
HSHichem Snoussi
Hichem Snoussi
Recently deepfake detection research focused on distinguishing fake faces from real ones when they are evaluated on test datasets similar to the training set. However, these approaches are proved to fail once test sets are different from training sets, such as forgeries created from unseen generation methods. Due to the distributional differences brought about by various forgery methods. It is challenging for current deepfake detectors to have the ability to perform well on cross-domain forgeries. In this paper, we introduce a generalized method designed for the cross-domain deep fake detection task. Our key idea is modifying Efficient-Net with cross-attention block and Siamese training to improve the generalization of detectors in cross-domain datasets. We investigate how the triple loss function effect model's generalization ability on a theoretical level. The AST network can balance the model's generalizability across domains. The detection ability of the network is improved by an intra-class compact loss.

19 October 2020 in Shock and Vibration
Verified
Journal Article
Open access
PDF file
TWTian Wang
Tian Wang
PXPing Xi
Ping Xi
BHBifu Hu
Bifu Hu
Product modeling has been applied in product engineering with success for geometric representation. With the application of multidisciplinary analysis, application-driven models need specific knowledge and time-consuming adjustment work based on the geometric model. This paper proposes a novel modeling technology named computer-aided design-supporting-simulation (CADSS) to generate multiphysics domain models to support multidisciplinary design optimization processes. Multiphysics model representation was analyzed to verify gaps among different domain models parameters. Therefore, multiphysics domain model architecture was integrated by optimization model, design model, and simulation model in consideration of domain models parameters. Besides, CADSS uses requirement space, domain knowledge, and software technology to describe the multidisciplinary models parameters and its transition. Depending on the domain requirements, the CADSS system extracts the required knowledge by decomposing product functions and then embeds the domain knowledge into functional features using software technology. This research aims to effectively complete the design cycle and improve the design quality by providing a consistent and concurrent modeling environment to generate an adaptable model for multiphysics simulation. This system is demonstrated by modeling turbine blade design with multiphysics simulations including computational fluid dynamics (CFD), conjugate heat transfer (CHT), and finite element analysis (FEA). Moreover, the blade multiphysics simulation model is validated by the optimization design of the film hole. The results show that the high-fidelity multiphysics simulation model generated through CADSS can be adapted to subsequent simulations.

12 November 2017
Verified
Book Chapter
Closed access
PDF file
TWTian Wang
Tian Wang
BHBifu Hu
Bifu Hu
PXPing Xi
Ping Xi
JLJixing Li
Jixing Li


27 August 2023
Verified
Conference Paper
Closed access
JCjunzhi chen
junzhi chen
TWTian Wang
Tian Wang
DLDeyuan Liu
Deyuan Liu
JWJian Wang
Jian Wang
YFYao Fu
Yao Fu
HSHichem Soussi
Hichem Soussi
Although deep learning has achieved high performance on many tasks, neural networks suffer from catastrophic forgetting in incremental learning scenarios. Class-Incremental Learning focuses on the scenario where new category data continuously arrives. PASS (Prototype Augmentation and Self-Supervised) is a classic method that does not require storing old exemplars. Based on PASS, we propose three improvement measures from the perspective of making the new and old tasks more balanced. Firstly, we propose a balanced prototype replay (Balanced PR) strategy to make the image number of new and old categories that the model head encounters more balanced. Secondly, at the feature knowledge distillation (FKD) stage, we assign different weights to the incoming data based on the confidence of the old model, which is named Focal FKD. The goal is to minimize interference with learning new task data while preserving the representation capability of the model backbone on old data. Thirdly, we combine FKD with weight regularization on the model backbone. In the feature space, FKD is beneficial for preserving the positions of the old categories while weight regularization helps to separate the new and old categories well. Experiments on the benchmark indicate that there has been obvious improvement in the performance of PASS after our modifications.