14 papers
Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models
Sultan Alshehri, Zhantao Yang, Han Zhang +1
Dual-encoder vision-language models (VLMs) expose a similarity interface that enables zero-shot retrieval but fails compositional constraints: queries like "umbrella and no person"…
PACE: Policy-Native Adaptive Decision Timing for Long-Horizon Reasoning
Chen Li, Zhantao Yang, Fangyi Chen +3
Long-horizon reasoning requires deciding not only what actions to take, but how many to execute open-loop before replanning. This number, the execution depth, balances replanning c…
Hierarchical SVG Tokenization: Learning Compact Visual Programs for Scalable Vector Graphics Modeling
Ximing Xing, Ziteng Xue, Zhenxi Li +8
Recent large language models have shifted SVG generation from differentiable rendering optimization to autoregressive program synthesis. However, existing approaches still rely on…
OmniCamera: A Unified Framework for Multi-task Video Generation with Arbitrary Camera Control
Yukun Wang, Ruihuang Li, Jiale Tao +7
Video fundamentally intertwines two crucial axes: the dynamic content of a scene and the camera motion through which it is observed. However, existing generation models often entan…
Addressing the ID-Matching Challenge in Long Video Captioning
Zhantao Yang, Huangji Wang, Ruili Feng +6
Generating captions for long and complex videos is both critical and challenging, with significant implications for the growing fields of text-to-video generation and multi-modal u…
MetaVLA: Unified Meta Co-training For Efficient Embodied Adaption
Chen Li, Zhantao Yang, Han Zhang +4
Vision-Language-Action (VLA) models show promise in embodied reasoning, yet remain far from true generalists-they often require task-specific fine-tuning, incur high compute costs,…
STELAR-VISION: Self-Topology-Aware Efficient Learning for Aligned Reasoning in Vision
Chen Li, Han Zhang, Zhantao Yang +4
Vision-language models (VLMs) have made significant strides in reasoning, yet they often struggle with complex multimodal tasks and tend to generate overly verbose outputs. A key l…
MAMBO-G: Magnitude-Aware Mitigation for Boosted Guidance
Shangwen Zhu, Qianyu Peng, Zhilei Shu +9
High-fidelity text-to-image and text-to-video generation typically relies on Classifier-Free Guidance (CFG), but achieving optimal results often demands computationally expensive s…
Instability in Diffusion ODEs: An Explanation for Inaccurate Image Reconstruction
Han Zhang, Jinghong Mao, Shangwen Zhu +6
Diffusion reconstruction plays a critical role in various applications such as image editing, restoration, and style transfer. In theory, the reconstruction should be simple - it j…
Accelerating Diffusion Sampling via Exploiting Local Transition Coherence
Shangwen Zhu, Han Zhang, Zhantao Yang +4
Text-based diffusion models have made significant breakthroughs in generating high-quality images and videos from textual descriptions. However, the lengthy sampling time of the de…
The Matrix: Infinite-Horizon World Generation with Real-Time Moving Control
Ruili Feng, Han Zhang, Zhantao Yang +7
We present The Matrix, the first foundational realistic world simulator capable of generating continuous 720p high-fidelity real-scene video streams with real-time, responsive cont…
Hierarchical Knowledge Graph Construction from Images for Scalable E-Commerce
Zhantao Yang, Han Zhang, Fangyi Chen +2
Knowledge Graph (KG) is playing an increasingly important role in various AI systems. For e-commerce, an efficient and low-cost automated knowledge graph construction method is the…
BACON: Improving Clarity of Image Captions via Bag-of-Concept Graphs
Zhantao Yang, Ruili Feng, Keyu Yan +13
Advancements in large Vision-Language Models have brought precise, accurate image captioning, vital for advancing multi-modal image understanding and processing. Yet these captions…
RTGen: Generating Region-Text Pairs for Open-Vocabulary Object Detection
Fangyi Chen, Han Zhang, Zhantao Yang +3
Open-vocabulary object detection (OVD) requires solid modeling of the region-semantic relationship, which could be learned from massive region-text pairs. However, such data is lim…