6 papers · 1 filter
PROBE: Manipulation-Grounded Visual Question Answering with VLM Agents
Vineet Bhat, Siyi Chen, Alex Zook +4
Vision-language Models (VLMs) excel at 2D grounding, spatial reasoning and agentic tool-based planning in static scenes. However, consider asking a home robot "Is my medication sti…
Data-Forcing Distillation: Restoring Diversity and Fidelity in Few-Step Video Generation
Siyi Chen, Shaowei Liu, Yixuan Jia +4
Recent progress has shown promise in distilling multi-step video diffusion models into efficient few-step students. Among them, Distribution Matching Distillation (DMD) and its suc…
On the Limits of Token Reduction for Efficient Unified Vision Language Training
Siyi Chen, Weiming Zhuang, Jingtao Li +1
Unified vision-language models (VLMs) integrate visual understanding and visual generation within a single autoregressive backbone, but their joint training is computationally expe…
SpaceTools: Tool-Augmented Spatial Reasoning via Double Interactive RL
Siyi Chen, Mikaela Angelina Uy, Chan Hee Song +6
Vision Language Models (VLMs) demonstrate strong qualitative visual understanding, but struggle with metrically precise spatial reasoning required for embodied applications. The ag…
Unfolding Videos Dynamics via Taylor Expansion
Siyi Chen, Minkyu Choi, Zesen Zhao +3
Taking inspiration from physical motion, we present a new self-supervised dynamics learning strategy for videos: Video Time-Differentiation for Instance Discrimination (ViDiDi). Vi…
Exploring Low-Dimensional Subspaces in Diffusion Models for Controllable Image Editing
Siyi Chen, Huijie Zhang, Minzhe Guo +3
Recently, diffusion models have emerged as a powerful class of generative models. Despite their success, there is still limited understanding of their semantic spaces. This makes i…