6 papers
How and What to Imagine? Visual Thinking in Unified Multimodal Models for Cross-View Spatial Reasoning
Qian Yang, Ankur Sikarwar, Huy Le +4
Cross-view spatial reasoning remains a weak spot for vision-language models (VLMs): they often reason in language and lose the fine-grained geometry needed for the task. Thinking w…
SlotVLA: Towards Modeling of Object-Relation Representations in Robotic Manipulation
Taisei Hanyu, Nhat Chung, Huy Le +10
Inspired by how humans reason over discrete objects and their relationships, we explore whether compact object-centric and object-relation representations can form a foundation for…
UNO: Unifying One-stage Video Scene Graph Generation via Object-Centric Visual Representation Learning
Huy Le, Nhat Chung, Tung Kieu +2
Video Scene Graph Generation (VidSGG) aims to represent dynamic visual content by detecting objects and modeling their temporal interactions as structured graphs. Prior studies typ…
Rethinking Progression of Memory State in Robotic Manipulation: An Object-Centric Perspective
Nhat Chung, Taisei Hanyu, Toan Nguyen +9
As embodied agents operate in increasingly complex environments, the ability to perceive, track, and reason about individual object instances over time becomes essential, especiall…
With Limited Data for Multimodal Alignment, Let the STRUCTURE Guide You
Fabian Gröger, Shuo Wen, Huyen Le +1
Multimodal models have demonstrated powerful capabilities in complex tasks requiring multimodal alignment, including zero-shot classification and cross-modal retrieval. However, ex…
BiMa: Towards Biases Mitigation for Text-Video Retrieval via Scene Element Guidance
Huy Le, Nhat Chung, Tung Kieu +2
Text-video retrieval (TVR) systems often suffer from visual-linguistic biases present in datasets, which cause pre-trained vision-language models to overlook key details. To addres…