A5063134138- Emotion and Mood Recognition
- Advanced Image Processing Techniques
- Speech and Audio Processing
- Image and Signal Denoising Methods
- Anomaly Detection Techniques and Applications
- Advanced Vision and Imaging
- Face recognition and analysis
- Analytical Chemistry and Sensors
- Advanced Optical Imaging Technologies
- Human Pose and Action Recognition
- Infrastructure Maintenance and Monitoring
- Industrial Vision Systems and Defect Detection
- Advanced Image Fusion Techniques
- Cognitive Abilities and Testing
- EEG and Brain-Computer Interfaces
- Surface Roughness and Optical Measurements
- Advanced Fiber Optic Sensors
- Image Processing Techniques and Applications
- Face and Expression Recognition
- Network Security and Intrusion Detection
- Social Robot Interaction and HRI
- Optical Coherence Tomography Applications
- Polydiacetylene-based materials and applications
- Blind Source Separation Techniques
- Data Stream Mining Techniques
Fudan University
2023-2024
The in-the-wild dynamic facial expression recognition (DFER) has been challenging due to several high-dynamics factors such as limited expression-related frames and variable non-expression noise in sequences. To provide more clips for DFER models, we propose a novel interpretable frequency-based method (Freq-HD) affective clip selection. It can select containing pure changes from sequences aid different network structures recognizing expressions accurately efficiently. We first design...
Automatic facial expression recognition (FER) based on face images is essential for affective robots, which are designed interactive companions and intelligent healthcare. Although existing DL-based FERs have made significant progress, an accurate FER model in robots challenging due to the subtle differences expressions across various scenarios. To address this issue, we propose a multigranularity region relation representation network (MGR <sup xmlns:mml="http://www.w3.org/1998/Math/MathML"...
The reason for wider color gamut of laser display systems rests on RGB lights whose coordinates very approach the locus chart, which means most saturated colors. A novel, practical and simplest method to reduce speckle is developed. digital cinema projector (LDCP) manufactured first time in world. LDCP enjoys all benefits 166% NTSC expanded gamut, beyond 9 024 ANSI lumens light output, a minimum sequential contrast ratio 2595:1, 2048 X 1080 resolution, etc.
Intelligent detection of product surface defects in the industrial scene is key to ensuring quality. On general benchmarks, current unsupervised anomaly techniques have achieved significant success. When used complex environments (e.g., large components with small defects), model needs be able adapt different imaging scenarios illumination and resolution) accurately detect localize anomalies, but its performance still far from satisfactory. Besides, unstable optical lighting environment for...
In the field of affective computing, fully leveraging information from a variety sensory modalities is essential for comprehensive understanding and processing human emotions. Inspired by process through which brain handles emotions theory cross-modal plasticity, we propose UMBEnet, brain-like unified modal network. The primary design UMBEnet includes Dual-Stream (DS) structure that fuses inherent prompts with Prompt Pool Sparse Feature Fusion (SFF) module. aimed at integrating different...
The acquisition and processing of visual data in industrial environments are paramount importance. High-resolution (HR) images offer superior clarity richer textural detail compared to low-resolution (LR) images. On the one hand, owing incorporation information, HR demonstrate substantially enhanced performance LR downstream applications, such as anomaly detection. other they provide valuable insights designers quality inspectors who require a detailed understanding Currently, majority...