A5101749121- Stochastic Gradient Optimization Techniques
- Generative Adversarial Networks and Image Synthesis
- Statistical Methods and Inference
- Network Packet Processing and Optimization
- 3D Shape Modeling and Analysis
- Neural Networks and Applications
- Algorithms and Data Compression
- Advanced Image Processing Techniques
- Biometric Identification and Security
- Face and Expression Recognition
- Privacy-Preserving Technologies in Data
- Domain Adaptation and Few-Shot Learning
- Advanced Vision and Imaging
- Service-Oriented Architecture and Web Services
- Machine Learning and Algorithms
- Forensic Fingerprint Detection Methods
- Human Motion and Animation
- Face recognition and analysis
- Sparse and Compressive Sensing Techniques
- Image and Signal Denoising Methods
- Fault Detection and Control Systems
- Advanced Statistical Methods and Models
- Network Security and Intrusion Detection
- Model Reduction and Neural Networks
- Data Analysis with R
Alibaba Group (China)
2022-2024
Henan University of Technology
2024
Hong Kong University of Science and Technology
2024
University of Hong Kong
2024
West China Hospital of Sichuan University
2023
Sichuan University
2023
Alibaba Group (Cayman Islands)
2022
Shanghai International Studies University
2011
Henan University
2010
University of California, Berkeley
2007
In this paper, we present a general-purpose solution to cartoon image synthesis with unpaired training data. contrast previous works learning pre-defined styles for specified usage scenarios (portrait or scene), aim train common translator which can not only simultaneously render exaggerated anime faces and realistic scenes, but also provide flexible user controls desired styles. It is challenging due the complexity of task absence paired The core idea proposed method introduce gated cycle...
This paper introduces DCT-Net, a novel image translation architecture for few-shot portrait stylization. Given limited style exemplars (~100), the new can produce high-quality transfer results with advanced ability to synthesize high-fidelity contents and strong generality handle complicated scenes (e.g., occlusions accessories). Moreover, it enables full-body via one elegant evaluation network trained by partial observations (i.e., stylized heads). Few-shot learning based is challenging...
Robust scatter estimation is a fundamental task in statistics. The recent discovery on the connection between robust and generative adversarial nets (GANs) by Gao et al. (2018) suggests that it possible to compute depth-like estimators using similar techniques optimize GANs. In this paper, we introduce general learning via classification framework based notion of proper scoring rules. This allows us understand both matrix depth function various GANs through lens variational approximations...
In this paper, based on the research of existing heterogeneous database integration systems, according to data exchange and sharing needs enterprise databases, a framework for system is proposed designed, key technologies implementation process are also described in detail. The provides middle platform achieve transparent operation seamless data.
Synthesis of high resolution images using Generative Adversarial Networks (GANs) is challenging, which usually requires numbers high-end graphic cards with large memory and long time training. In this paper, we propose a two-stage framework to accelerate the training process synthesizing images. High are first transformed small codes via trained encoder decoder networks. The code in latent space times smaller than original Then, train generation network learn distribution codes. way,...
Body reshaping is an important procedure in portrait photo retouching. Due to the complicated structure and multifarious appearance of human bodies, existing methods either fall back on 3D domain via body morphable model or resort keypoint-based image deformation, leading inefficiency unsatisfied visual quality. In this paper, we address these limitations by formulating end-to-end flow generation architecture under guidance structural priors, including skeletons Part Affinity Fields, achieve...