A5008172477- Space Satellite Systems and Control
- Generative Adversarial Networks and Image Synthesis
- Advanced Vision and Imaging
- Computer Graphics and Visualization Techniques
- Muscle activation and electromyography studies
- Stroke Rehabilitation and Recovery
- Prosthetics and Rehabilitation Robotics
- Robotics and Sensor-Based Localization
- Space exploration and regulation
- Digital Rights Management and Security
- Water Quality Monitoring and Analysis
- Innovation in Digital Healthcare Systems
- Water Quality Monitoring Technologies
- Wastewater Treatment and Nitrogen Removal
- Spaceflight effects on biology
- Morphological variations and asymmetry
- Digital Image Processing Techniques
- Image Retrieval and Classification Techniques
- Cinema and Media Studies
Dalian University of Technology
2023-2025
Google (United States)
2024
Shenyang University of Technology
2021-2022
We introduce an Extended Textual Conditioning space in text-to-image models, referred to as $P+$. This consists of multiple textual conditions, derived from per-layer prompts, each corresponding a layer the denoising U-net diffusion model. show that extended provides greater disentangling and control over image synthesis. further Inversion (XTI), where images are inverted into $P+$, represented by tokens. XTI is more expressive precise, converges faster than original (TI) space. The...
Recent advances in generative imagery have brought forth outpainting and inpainting models that can produce high-quality, plausible image content unknown regions. However, the these hallucinate is necessarily inauthentic, since they are unaware of true scene. In this work, we propose RealFill, a novel approach for completion fills missing regions an with should been there. RealFill model personalized using only few reference images These do not to be aligned target image, taken drastically...
Wearable exoskeletons can help people with mobility impairments by improving their rehabilitation. As electromyography (EMG) signals occur before movement, they be used as input for the to predict body’s movement intention. In this paper, OpenSim software is determine muscle sites measured, i.e., rectus femoris, vastus lateralis, semitendinosus, biceps lateral gastrocnemius, and tibial anterior. The surface (sEMG) inertial data are collected from lower limbs while human body walking, going...
The increasing of satellite launch missions, collisions and disintegration events space objects, the vast number near-Earth asteroids pose significant challenges to traffic management. Space situational awareness aims detect, track, classify, characterize as well assess events, playing a crucial role in application artificial intelligence has unique advantages, especially enhancing comprehensive abilities such knowledge, analysis, decision-making. Its outstanding transfer learning lifelong...
In view of the complexity actual wastewater treatment process, it is very difficult to establish an accurate mechanism model. The simple model cannot accurately describe all characteristics and moreover complex relationship between variables be completely included in general, prediction accuracy not high. Therefore, data-driven modeling method used actural process data. this paper, hybrid activated sludge system established, which combines simplified BSM1 with BP neural network. model,...
Recent advances in generative imagery have brought forth outpainting and inpainting models that can produce high-quality, plausible image content unknown regions. However, the these hallucinate is necessarily inauthentic, since they are unaware of true scene. In this work, we propose RealFill, a novel approach for completion fills missing regions an with should been there. RealFill model personalized using only few reference images These do not to be aligned target image, taken drastically...
In this paper, the link angle trajectory tracking of a Lower Limb Rehabilitation Robot (LLRR) under passive training stage has been investigated. To deal with mismatch between model and actual controlled object, control method based on RTO (Real Time Optimization) enforcing SCFO (Sufficient Condition for Feasibility Optimality) LLRR is proposed, in which use used to ensure that results have desirable feasibility. The proposed verified by simulation experiments model. show can accurately...