OS-ATLAS: A Foundation Action Model for Generalist GUI Agents

PuLID: Pure and Lightning ID Customization via Contrastive Alignment

Paint by Inpaint: Learning to Add Image Objects by Removing Them First

Automatic Creative Selection with Cross-Modal Matching

STT: Stateful Tracking with Transformers for Autonomous Driving

Data-Efficient Multimodal Fusion on a Single GPU

SAGS: Structure-Aware 3D Gaussian Splatting

Genie: Generative Interactive Environments

GPT-4V(ision) system card

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

MIME: Human-Aware 3D Scene Generation

Recognize Anything: A Strong Image Tagging Model

CAPE: Camera View Position Embedding for Multi-View 3D Object Detection

Improving Factuality and Reasoning in Language Models through Multiagent Debate

ORCa: Glossy Objects as Radiance Field Cameras

Random-Access Neural Compression of Material Textures

Scaling Vision Transformers to 22 Billion Parameters

DINOv2: Learning Robust Visual Features without Supervision

StereoDistill: Pick the Cream from LiDAR for Distilling Stereo-based 3D Object Detection

A Good Prompt Is Worth Millions of Parameters? Low-resource...