自動音声認識 (ASR)、大規模言語モデル (LLM)、スマートグラスを組み合わせた「インタラクティブサイクルモデル」は、人間とコンピュータの自然でシームレスなインタラクションを実現する可能性を秘めている。
This case study explores the capabilities and limitations of Claude 3.5 Computer Use, a GUI agent, in automating diverse desktop tasks, highlighting its strengths in web search, workflow, and productivity applications while revealing challenges in handling dynamic interfaces and scrolling-based navigation.
AI 기반 피드백 루프가 사용자의 행동 변화, 동기 부여, 웰빙에 미치는 긍정적, 부정적 영향을 동시에 살펴보고, 기술 발전과 사용자 웰빙 사이의 균형점을 제시합니다.
SimTube is a novel AI system that simulates diverse and believable audience comments on videos before their release, leveraging multimodal data analysis and user personas to provide valuable feedback for content creators.
Aligning large language models (LLMs) with expert-crafted dialogue scripts, particularly through a prompting-based approach called Script-Aligned Generation (SAG), significantly improves the performance of psychotherapy chatbots in terms of linguistic quality, therapeutic relevance, engagement, and user motivation compared to rule-based and pure LLM chatbots.
CoPrompter is a new tool that helps prompt engineers identify and fix misalignments between their instructions and the output of large language models (LLMs), leading to more effective prompt design.
LLM 기반 가상 에이전트의 성격 특성(페르소나)은 사용자의 사회적 평가, 감정적 경험, 행동적 참여에 큰 영향을 미치며, 특히 외향적인 페르소나는 더 긍정적인 반응을 이끌어냅니다.
Sensory perception acts as a universal language that bridges communication gaps between disciplines, particularly in media art, HCI, and AI, by translating abstract concepts into tangible experiences.
GazeGen is a novel system that leverages real-time gaze estimation to enable intuitive and efficient visual content generation and editing, enhancing user experience and accessibility in augmented reality environments.
깊은 사랑에 빠졌을 때 느끼는 감사는 단순한 감정을 넘어 관계의 본질을 드러내는 표현이다.