Chen Jingdong
Technical Director of the Foundation Intelligence Department at Ant Group; Senior Algorithm Expert
Chen Jingdong is a Technical Director of the Foundation Intelligence Department at Ant Group and a senior algorithm expert, leading R&D in multimodal foundation models. His OCR-based intelligent claims processing project and remote-sensing–based rural credit initiative received the CCF Science and Technology Progress Second Prize in 2022 and 2023, respectively. He has published numerous papers in computer vision and machine learning, with over 5,300 citations. His achievements in speech recognition were recognized as one of MIT Technology Review’s Top 10 Breakthrough Technologies in 2016.
Topic
Multimodal Large Model R&D at Ant Group
his talk will introduce Ant Group’s R&D practices and technical explorations in multimodal large models, with a focus on the Ling-VL and Ming-Omni series. It will cover the evolution of Bailing’s multimodal model technology, from unified audio, vision, image, and text capabilities to native multimodal training for real-world tasks and comprehensive optimization of multimodal agent capabilities. The talk will cover: The open-source Bailing multimodal model series and the evolution of its technical approaches Ling-VL: Native multimodal models designed for real-world tasks Design Assistant: Image generation for design applications Advances in speech capabilities, including speech understanding, speech generation, and end-to-end speech interaction models Technical innovations in the Ming-Omni series
