Search UW

On this page

2026

  1. NeurIPS 2026 Oral

    Language-Conditioned World Modeling for Visual Navigation, by Yifei Dong, Fengyi Wu, and collaborators, connects language instructions, imagined futures, and navigation actions.

  2. ECCV 2026 · UniWM

    UniWM, by Yifei Dong, Fengyi Wu, and collaborators, integrates memory, planning, and foresight for visual navigation.

  3. IROS 2026 · HA-VLN 2.0

    HA-VLN 2.0 is an open benchmark and leaderboard for human-aware navigation in discrete and continuous environments with dynamic multi-human interactions.

  4. Two USDOT-funded research projects

    Funding from the U.S. Department of Transportation through PacTrans supports two projects on 3D urban scene reconstruction and transportation asset mapping.

  5. Fengyi Wu receives a UW Graduate School Top Scholar Award

    Fengyi Wu received the 2026 UW GSFEI Top Scholar Award.

  6. Two papers at ACL 2026

    Sign-Language Datasets at Scale was accepted to the ACL 2026 main conference, and GoViG, on goal-conditioned navigation instruction generation, was accepted to Findings of ACL 2026.

  7. ICLR 2026 Oral

    Lossless Hierarchical Speculative Decoding is an ICLR 2026 oral paper. The method accelerates language-model inference while preserving the target model’s output distribution.

2025

  1. NeurIPS 2025 Oral

    MaxSup, with former visiting student Yuxuan Zhou, was accepted for an oral presentation at NeurIPS 2025.

  2. Carwein–Andrews Ph.D. Fellowship for Yifei Dong

    Yifei Dong received the Carwein-Andrews Fellowship.

  3. CVPR Anti-UAV Workshop · Best Paper Award

    Professor Cheng gave a keynote at the CVPR 2025 Anti-UAV Workshop. Securing the Skies received the workshop's Best Paper Award.

  4. Oral papers at AAAI and NAACL 2025

    VDAct and ProMQA were presented as oral papers at AAAI 2025 and NAACL 2025.

  5. Yifei Dong receives a UW Graduate School Top Scholar Award

    Yifei Dong received the 2025 UW GSFEI Top Scholar Award, a Graduate School award for incoming doctoral students.

  6. Four papers across CVPR 2025 and its workshops

    StableAnimator, on identity-preserving image animation, and Emphasizing Discriminative Features, on dataset distillation in complex scenarios, appeared at CVPR 2025. Securing the Skies and Why We Feel appeared in CVPR workshops.

  7. ICLR 2025 · MetaDesigner

    MetaDesigner was accepted to ICLR 2025. The project studies controllable, user-centered multilingual WordArt synthesis.