Qifu Technology, in collaboration with Beijing Jiaotong University, announced that their paper, "Leveraging MLLM Embeddings and Attribute Smoothing for Compositional Zero-Shot Learning," has been accepted by the International Joint Conference on Artificial Intelligence (IJCAI) 2025. IJCAI is recognized as a premier international conference in artificial intelligence, with an acceptance rate of 19.3%.
The research introduces an innovative framework named TRIDENT, which integrates Multimodal Large Language Model (MLLM) embeddings and attribute smoothing. This framework aims to identify novel combinations of attributes and objects using existing knowledge, overcoming challenges such as background interference and limited semantic capture. TRIDENT has demonstrated state-of-the-art performance across multiple datasets.
In the fintech sector, TRIDENT shows potential for intelligent risk control by analyzing multimodal data to detect emerging fraud patterns faster and improve assessment accuracy. It also enables more precise understanding of complex user inquiries in customer service, providing personalized and efficient support. This achievement highlights Qifu Technology's commitment to AI innovation and R&D investment.
The content on BeyondSPX is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.