lightbulb现实痛点人机协同标注与示例选择领域专家标注图像数据成本高、效率低且规则难以解释。IUI '25HEPHA: A Mixed-Initiative Image Labeling Tool for Specialized Domains
lightbulb现实痛点人机协同标注与示例选择文本到SQL模型在新领域使用中因数据不足表现不佳。IUI '25Text-to-SQL Domain Adaptation via Human-LLM Collaborative Data Annotation
lightbulb现实痛点人机协同标注与示例选择用户在复杂或主观任务的标注中难以理解模型状态和数据边界。CHI '25Supporting Co-Adaptive Machine Teaching through Human Concept Learning and Cognitive Theories
lightbulb现实痛点人机协同标注与示例选择当前智能助手缺乏通用能力、安全性和隐私保护,难以高效操作各种手机应用。CHI '25AppAgent: Multimodal Agents as Smartphone Users
lightbulb现实痛点人机协同标注与示例选择普通用户在缺乏标准数据时难以优化提示以进行高质量数据标注。CHI '25Prompting in the Dark: Assessing Human Performance in Prompt Engineering for Data Labeling When Gold Labels Are Absent
lightbulb现实痛点人机协同标注与示例选择人类在标注机器人轨迹偏好时常感到认知负担重,标注一致性差。IUI '24FARPLS: A Feature-Augmented Robot Trajectory Preference Labeling System to Assist Human Labelers’ Preference Elicitation
lightbulb现实痛点人机协同标注与示例选择灾害管理中,新手标注的数据常与专家存在较大差异,影响AI模型质量。IUI '24Closing the Knowledge Gap in Designing Data Annotation Interfaces for AI-powered Disaster Management Analytic Systems