Computer Vision

Orient Anything: Learning Robust Object Orientation Estimation from Rendering 3D Models

DiTCtrl: Exploring Attention Control in Multi-Modal Diffusion Transformer for Tuning-Free Multi-Prompt Longer Video Generation

DepthLab: From Partial to Complete

3DGraphLLM: Combining Semantic Graphs and Large Language Models for 3D Scene Understanding

1.58-bit FLUX

VidTwin: Video VAE with Decoupled Structure and Dynamics

Large Motion Video Autoencoding with Cross-modal Video VAE

Distilled Decoding 1: One-step Sampling of Image Auto-regressive Models with Flow Matching

Toward Robust Hyper-Detailed Image Captioning: A Multiagent Approach and Dual Evaluation Metrics for Factuality and Coverage

MotiF: Making Text Count in Image Animation with Motion Focal Loss

CLEAR: Conv-Like Linearization Revs Pre-Trained Diffusion Transformers Up

UIP2P: Unsupervised Instruction-based Image Editing via Cycle Edit Consistency

Parallelized Autoregressive Visual Generation

LeviTor: 3D Trajectory Oriented Image-to-Video Synthesis

IDOL: Instant Photorealistic 3D Human Creation from a Single Image

DI-PCG: Diffusion-based Efficient Inverse Procedural Content Generation for High-quality 3D Asset Creation

Affordance-Aware Object Insertion via Mask-Aware Dual Diffusion

Thinking in Space: How Multimodal Large Language Models See, Remember, and Recall Spaces

Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation

PixelMan: Consistent Object Editing with Diffusion Models via Pixel Manipulation and Generation