SBIR/STTR Award attributes
In a business where collaboration and rapport is critical, having Hokkien speakers may make or break a relationship with a partner nation. The military has linguists and contract translators, in addition to organic foreign language speakers; however, resources are extremely limited, so it is critical to invest in technologies that can enhance, enable, and complement existing resources. To address this problem, Black Cape would like to build a capability called Sid-ML, starting with a SBIR Phase I feasibility study to ascertain the requirements and technology needed to create a self-contained, speech-to-speech translator for low-density languages, starting with Chinese Hokkien. Low-density language speech-to-speech translation is incredibly difficult; however, Black Cape is experienced in natural language processing and large language models. For the Phase I feasibility study, Black Cape will focus on five key areas: artificial intelligence (AI) and machine learning (ML) models; required speech and text datasets; system and model microservice integration; commercial, off-the-shelf (COTS) handheld hardware to facilitate system and model integration; and integration roadmap with other relevant systems and capabilities.
