Recently, Wang Yongtao’s research team at the Wang Xuan Institute of Computer Technology at Peking University has made significant strides in the field of intelligent driving. Concentrating on crucial areas such as multimodal fusion, open-world perception, high-level end-to-end VLA (Vision-Language-Action) intelligent driving systems, and the synthesis and reconstruction of driving scene data, the team has sequentially unveiled a series of innovative works. These include R4Det, VL-SAM v2, AutoOcc, OpenAD, KnowVal, and DrivingGaussian++. Their achievements have been published in or accepted by leading international conferences in artificial intelligence, such as ICCV 2025, NeurIPS 2025, and CVPR 2026, as well as the prestigious international journal TPAMI. Furthermore, the team has applied for multiple invention patents.
