Sub-Project 9(SungHyun Sim): Development of XAI-based smart safety detection technology for children’s daily life safety 

Key Research Activities

• Develop an Interpretable Multimodal Attention Layer for the integrated learning of video and audio data
• Develop a Contrastive Language-Image-Voice Pre-training (CLIVP) embedding methodology to link child monitoring data with explainable, hyper-scale AI
• Develop a hyper-scale behavior model and linkage system specialized for child living environments, utilizing monitoring data and explainable, hyper-scale AI

Research Description

• Develop an interpretable multimodal attention layer for integrated learning of image and audio data
• Develop a Contrastive Language-Image-Voice Pre-training (CLIVP) embedding methodology to connect children’s daily life monitoring data with explainable large-scale AI
• Develop a large-scale, behavior-specialized model and integration system for children's living environments using children’s daily life monitoring data and explainable large-scale AI