Ranjay Krishna

Innovative method simplifies training models for complex image classification.

A new benchmark assesses how machines plan complex tasks with different data types.

A new training method enhances the compositionality of vision-language models.

EVE simplifies robot training using augmented reality for everyday users.

A new benchmark reveals gaps in visual understanding of large language models.

A new method enhances image descriptions for training AI models.

Including non-English data improves vision-language model performance and cultural understanding.

A new framework enhances reasoning in language models through visual sketches.

A new method improves how AI models interpret spatial and temporal relationships.

Discover how RONAR helps robots explain their actions in simple terms.

OneDiffusion transforms text into images, enhancing creativity for everyone.

Perception Tokens enhance AI's ability to understand and interpret images.

Learn how Negative Token Merging is changing AI image generation.

A new approach improves machine spatial reasoning for real-world applications.

A new method to evaluate AI's image and video generation using scene graphs.