The Rapid-Rich Object Search (ROSE) Lab is a joint collaboration between NTU and Peking University (PKU), drawing upon each university’s strength in media, computer vision, and image processing technology. It aims to build one of the largest database of structured domain objects in Asia, and to develop visual object search technology on media cloud platforms to support real-time applications with scalability to user demand. With the proliferation of mobile devices, internet searches are increasingly performed on tablets and smartphones. Current mobile search is mainly based on text, with limited capability in describing real-world objects. In many situations, a picture taken with a mobile device is simply more descriptive. We believe visual object search will be the trend in the coming years and create huge economic value and opportunities.
Centre activities and achievements
The ROSE Lab’s research focuses on three areas:
1. Large-scale structured database of media objects
2. Scalable visual object search
3. Media cloud platform
To date, the ROSE Lab has developed 4 products, including the visual object search system and the Image Recognition module for Tencent’s WeChat Smart Platform. There also have several technical disclosures been generated by the ROSE Lab. The ROSE Lab has received $7.1 million in third-party co-funding commitments and secured 8 partners, including Tencent, one of the largest Internet providers in Asia, and NVIDIA, the world’s leading visual computing company.
Our lab's objectives are well aligned with Singapore’s goal of becoming a global media capital by promoting research into visual object search technology. The collaborative efforts with worldwide partners of ROSE Lab will lead to key advancements in new generation rapid and rich visual and object search technology, as well as bring new economic opportunities and benefits.
Schematic diagram of the Visual Object Search Technology