- Face recognition and analysis
- Face and Expression Recognition
- Digital Media Forensic Detection
- Emotion and Mood Recognition
- Catalytic Processes in Materials Science
- Simulation and Modeling Applications
- Catalysis and Hydrodesulfurization Studies
- Advanced Steganography and Watermarking Techniques
- Biometric Identification and Security
- Image Processing Techniques and Applications
- 3D Shape Modeling and Analysis
- Zeolite Catalysis and Synthesis
- Vehicle emissions and performance
- Law in Society and Culture
- Archaeological Research and Protection
- Biodiesel Production and Applications
- Advanced Computing and Algorithms
- Advanced Combustion Engine Technologies
- Industrial Technology and Control Systems
- Fuel Cells and Related Materials
- Advanced Image Processing Techniques
- Generative Adversarial Networks and Image Synthesis
- Brain Tumor Detection and Classification
- Manufacturing Process and Optimization
- Industrial Gas Emission Control
Ping An (China)
2020-2024
Hunan University
2024
Shandong Institute of Commerce & Technology
2021
National Engineering Research Center for Wheat
2021
University of Jinan
2019
Source camera identification has long been a hot topic in the field of image forensics. Besides conventional feature engineering algorithms developed based on studying traces left upon shooting, several deep-learning-based methods have also emerged recently. However, performance is susceptible to content and far from satisfactory for small patches real demanding applications. In this paper, an efficient patch-level source method proposed convolutional neural network. First, order obtain...
To quickly construct the orthopedic plates and to conveniently edit it, a novel method for designing is put forward based on feature idea parameterization. Firstly, attached existing or repaired bone mod... | Find, read cite all research you need Tech Science Press
Camera model identification is a hot topic in the field of image forensics. In this paper, patch-level camera method based on convolutional neural network proposed. Firstly, inspired by pre-processing traditional method, an automatic residual extraction module designed order to avoid subsequent being affected content. Secondly, modified SqueezeNet proposed extract related features within patches. Finally, effectiveness verified under strict evaluation protocol, which largest public forensic...
Sensor Pattern Noise (SPN) has proven to be an effective fingerprint for source camera identification, while its estimation accuracy heavily relies on denoising algorithm. In this paper, identification scheme based Multi-Scale Expected Patch Log Likelihood (MSEPLL) algorithm is proposed, firstly. With enhanced prior modeling across multiple scales, MSEPLL can accurately restore the original image. As a consequence, estimated SPN less influenced by image content. Secondly, problem formulated...
Facial expression is an essential factor in conveying human emotional states and intentions. Although remarkable advancement has been made facial recognition (FER) task, challenges due to large variations of patterns unavoidable data uncertainties still remain. In this paper, we propose mid-level representation enhancement (MRE) graph embedded uncertainty suppressing (GUS) addressing these issues. On one hand, MRE introduced avoid learning being dominated by a limited number highly...
Facial expression is an essential factor in conveying human emotional states and intentions. A common strategy used for facial recognition (FER) encoding representations from images. Although remarkable advancement has been made, challenges due to large variations of patterns unavoidable hard samples still remain. In this paper, we propose dual-level representation enhancements (DLRE) addressing these issues. On one hand, mid-level enhancement (MRE) introduced avoid learning being dominated...
Based on convolutional neural network (CNN), the problem of robust patch level camera model identification is studied in this paper. Firstly, an effective feature representation proposed by concatenating a multiscale residual prediction module as well original RGB images. Motivated exploration multi-scale characteristic, automatically learn images to avoid subsequent CNN being affected scene content. Color channel information integrated for enhanced diversity inputs. Secondly, modified...
Face anti-spoofing is an important problem for both academic research and industrial face recognition systems. Most of the existing methods take it as a classification task on individual static images, where motion pattern differences in consecutive real or fake sequences are ignored. In this work, we propose novel method to identify spoofing patterns using information. Different from previous methods, proposed makes decision disentangled feature level, based observation that features could...