New method improves object localization using relationships between language and images.
― 6 min read
Boqing Gong
New methods improve machines' ability to create images from textual prompts.
― 5 min read
New method improves video captioning using image-language models.
― 6 min read
VideoPrism helps interpret and analyze video content effectively.
― 5 min read
Research reveals how trigger patches influence image generation in diffusion models.
― 6 min read
A new approach to improve text-to-image model prompts for enhanced results.
― 5 min read
SOAR improves action recognition accuracy in drone footage analysis.
― 5 min read
Introducing Long Video Masked Autoencoders for better video understanding.
― 6 min read
HypDAE transforms how we create images from minimal examples.
― 6 min read
DAVE dataset captures complex road scenarios for better AI training.
― 7 min read