Japan, Oct. 13 -- NEC Corporation has developed two new technologies for Vision-Language Models (VLMs), VLM-based Local Region Understanding and Continual Learning for VLMs, which are expected to be utilized in rescue operations and recovery efforts by local governments and other organizations. VLM-based Local Region Understanding accurately identifies small-scale damage visible within images collected from disaster sites. Continual Learning for VLMs enables new types of damage information to be added as search criteria while maintaining the VLM's original recognition capabilities. These technologies are the result of joint research with the Okatani Laboratory at Tohoku University in Japan.VLMs are expected to support rescue and recovery ef...