Qi Wu

AerialVLN improves drone navigation using language and visual data.

ClipVID improves object detection by focusing on unique identities across frames.

A simplified method improves efficiency in text-video matching.

A novel approach to assess image generation quality based on text descriptions.

Introducing a framework for better analysis of irregularly sampled time series data.

A new technique helps language models generate diverse outputs beyond text.

A new method enhances security of Vision Transformers against adversarial attacks.

ModaVerse simplifies how we transform and interpret various types of data.

NaVid helps robots follow human instructions using video, improving real-world navigation.

A new method improves CATE estimation and enhances decision-making in various fields.

G-NeRF innovates generating new views from single images with enhanced geometry techniques.

MotionLLM creates human movements from text for single and multi-person scenarios.

Data-parallel ANARI improves rendering efficiency and quality in scientific visualization.

Exploring how machines can follow human directions in real-world spaces.

Combining language understanding and vision enhances robot navigation capabilities.

XLIP enhances diagnosis by integrating medical images and text descriptions.

This article delves into the intriguing properties and production mechanisms of charmonium states.

This article discusses the safety and security issues in multimodal AI systems.

Learn about Legionella pneumonia, its risks, symptoms, and importance of early treatment.

Learn how new watermarking techniques protect digital art and creative ideas.

AbilityLens standardizes evaluation for multimodal large language models.

Hypernetworks transform data analysis, filling gaps and improving precision in dynamic simulations.

Research focuses on teaching machines to follow spoken and written navigation instructions.

A new way to render stunning visuals in real time.