Latest Articles for Objects

Categories are key structures for organizing mathematical objects and their relationships.

A new framework improves object visibility in complex images through innovative methods.

CUPID aids users in understanding generative models through visual analysis of generated images.

GenMM improves realistic insertion of 3D objects in videos and LiDAR scans.

A new approach improves image generation from sketches by separating object and scene creation.

A new method helps robots find fallen objects using sound.

OVExp combines language and vision for effective object navigation in varied environments.

A look into why smaller objects often feel heavier than larger ones.

Robots enhance task performance by understanding object interaction.

This method simplifies adding objects to images with text prompts, ensuring natural results.

Exploring the connection between gaze, attention, and object recognition.

A new task challenges models to analyze entire video streams for better comprehension.

A novel approach to track interactions without templates.

Robots learn tasks faster through in-context imitation learning.

A look into how mathematical structures shape our understanding of mathematics.

This article evaluates sentiment and meaning in image captions.

New method helps robots learn tasks by watching human demonstrations.

New methods enable machines to quickly identify unfamiliar objects with minimal input.

Learn how surroundings affect machines' ability to recognize objects in images.

New tool H-POPE improves accuracy of vision-language models.

Innovative method combines 2D images with depth to design realistic 3D spaces.

A new method improves 3D mapping efficiency and accuracy.

VeriGraph enhances robot cleaning efficiency through advanced scene graph technology.

GLOVER enables robots to grasp objects more effectively using language and visual cues.

Introducing GalaxyEdit, a dataset improving object editing in images.

Examining how we move our eyes and hands in daily tasks.

Robots now tilt objects for easier lifting, enhancing efficiency in various tasks.

DexGrip can hold, twist, and manipulate objects like a human hand.

Research highlights the challenge of verb understanding in multimodal AI models.

Discover how feature inversion reveals the inner workings of DETR networks.

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

Drones improve efficiency in farming by learning smarter flight paths for object detection.

Researchers enhance 3D reconstructions using synthetic data for better results.

A deep dive into circuit algebras and their real-world applications.