A5084894084- Human-Automation Interaction and Safety
- Robotics and Automated Systems
- Supercapacitor Materials and Fabrication
- Human Motion and Animation
- Visual Attention and Saliency Detection
- Advancements in Battery Materials
- Advanced Battery Materials and Technologies
- Speech and dialogue systems
- 3D Shape Modeling and Analysis
- Ethics and Social Impacts of AI
- Explainable Artificial Intelligence (XAI)
- Image Processing and 3D Reconstruction
- Context-Aware Activity Recognition Systems
- Virtual Reality Applications and Impacts
Xi'an Jiaotong University
2020-2024
Hyperbolic spaces allow for more efficient modeling of complex, hierarchical structures, which is particularly beneficial in tasks involving multi-modal data. Although hyperbolic geometries have been proven effective language-image pre-training, their capabilities to unify language, image, and 3D Point Cloud modalities are under-explored. We extend the modality contrastive pre-training. Additionally, we explore entailment, gap, alignment regularizers learning embeddings facilitating transfer...
Potassium-ion batteries have attracted increasing attention for next-generation energy storage systems due to their high density and abundance of potassium. However, the lack suitable anode highly hampers its practical application large ionic radius K+ . Herein, a Se3 P4 @mesoporous carbon (Se3 @C) composite is reported as high-performance potassium-ion batteries. The @C synthesized through an in situ combination reaction between red phosphorus Se within porous matrix. In this way,...
The existing work on task assignment of human-AI cooperation did not consider the differences between individual team members regarding their capabilities, leading to sub-optimal completion results. In this work, we propose a capability-aware shared mental model (CASMM) with components grouping and negotiation, which utilize tuples break down tasks into sets scenarios relating difficulties then dynamically merge ideas raised by human AI through negotiation. We implement prototype system...
To address critical challenges in effectively identifying user intent and forming relevant information presentations recommendations VR environments, we propose an innovative condition-based multi-modal human-AI cooperation framework. It highlights the tuples (intent, condition, prompt, action prompt) 2-Large-Language-Models (2-LLMs) architecture. This design, utilizes "condition" as core to describe tasks, dynamically match interactions with intentions, empower generations of various...
In this paper, we present a novel approach to advancing augmented reality (AR) dialogue systems, bridging the gap between two-dimensional spaces and immersive virtual environments. We construct "SIMMC2-Point" dataset, which transforms original SIMMC2 dataset from (VR) into AR environments, highlighting additional introduced pointing modality support understanding user's multi-modal intentions in AR. By harnessing power of BART CLIP models, design architecture dialogues that effectively...