A5070343486- Gaze Tracking and Assistive Technology
- Robotics and Sensor-Based Localization
- Emotion and Mood Recognition
- Video Surveillance and Tracking Methods
- Recommender Systems and Techniques
- Functional Brain Connectivity Studies
- Advanced Sensor and Energy Harvesting Materials
- Consumer Market Behavior and Pricing
- Interactive and Immersive Displays
- 3D Shape Modeling and Analysis
- Tactile and Sensory Interactions
- Customer churn and segmentation
- Robotic Path Planning Algorithms
- EEG and Brain-Computer Interfaces
- Image Processing and 3D Reconstruction
- Human Motion and Animation
- CCD and CMOS Imaging Sensors
Northwestern Polytechnical University
2016-2024
Graph neural networks (GNNs) have demonstrated efficient processing of graph-structured data, making them a promising method for electroencephalogram (EEG) emotion recognition. However, due to dynamic functional connectivity and nonlinear relationships between brain regions, representing EEG as graph data remains great challenge. To solve this problem, we proposed multi-domain based representation learning (MD
Abstract Maneuvering target tracking means that UAV observes the through sensors, follows and maintains range with target. Aiming at problem of UAV, we built motion model based on reinforcement learning, verified feasibility effectiveness method in simulation environment.
Recent diffusion models have demonstrated remarkable performance in both 3D scene generation and perception tasks. Nevertheless, existing methods typically separate these two processes, acting as a data augmenter to generate synthetic for downstream In this work, we propose OccScene, novel mutual learning paradigm that integrates fine-grained high-quality unified framework, achieving cross-task win-win effect. OccScene generates new consistent realistic scenes only depending on text prompts,...
In this paper, we present TAPTRv2, a Transformer-based approach built upon TAPTR for solving the Tracking Any Point (TAP) task. borrows designs from DEtection TRansformer (DETR) and formulates each tracking point as query, making it possible to leverage well-studied operations in DETR-like algorithms. TAPTRv2 improves by addressing critical issue regarding its reliance on cost-volume,which contaminates query\'s content feature negatively impacts both visibility prediction cost-volume...