Computer Vision

VideoGen-of-Thought (VGoT) creates high-quality, multi-shot videos by collaboratively generating scripts, keyframes, and video clips, ensuring narrative consistency and visual coherence.

SNOOPI supercharges one-step diffusion model distillation with enhanced guidance, achieving state-of-the-art performance by stabilizing training and enabling negative prompt control.

GSQ-GAN, a novel image tokenizer, achieves superior reconstruction quality with 16x downsampling using grouped spherical quantization, enabling efficient scaling for high-fidelity image generation.

OmniCreator: Self-supervised unified image+video generation & universal editing.

VideoLights: a novel framework for joint video highlight detection & moment retrieval, boosts performance via feature refinement, cross-modal & cross-task alignment, achieving state-of-the-art results…

TinyFusion, a novel learnable depth pruning method, crafts efficient shallow diffusion transformers with superior post-fine-tuning performance, achieving a 2x speedup with less than 7% of the original…

SWITTI: a novel scale-wise transformer achieves 7x faster text-to-image generation than state-of-the-art diffusion models, while maintaining competitive image quality.

Unified 3D latent representation (SLAT) enables versatile high-quality 3D asset generation, significantly outperforming existing methods.

PhysGame benchmark unveils video LLMs’ weaknesses in understanding physical commonsense from gameplay videos, prompting the creation of PhysVLM, a knowledge-enhanced model that outperforms existing mo…

One-shot image to realistic, animatable talking avatar! Novel pipeline uses diffusion models and a hybrid 3DGS-mesh representation, achieving seamless generalization and precise control.

NitroFusion achieves high-fidelity single-step image generation using a dynamic adversarial training approach with a specialized discriminator pool, dramatically improving speed and quality.

NegToMe: Image-based adversarial guidance improves image generation diversity and reduces similarity to copyrighted content without training, simply by using images instead of negative text prompts.

Presto: a novel video diffusion model generates 15-second, high-quality videos with unparalleled long-range coherence and rich content, achieved through a segmented cross-attention mechanism and the L…

VISTA synthesizes long-duration, high-resolution video instruction data, creating VISTA-400K and HRVideoBench to significantly boost video LMM performance.

DisCoRD: Rectified flow decodes discrete motion tokens into continuous, natural movement, balancing faithfulness and realism.

AlphaTablets: A novel 3D plane representation enabling accurate, consistent, and flexible 3D planar reconstruction from monocular videos, achieving state-of-the-art results.

RollingDepth: Achieving state-of-the-art video depth estimation without using complex video models, by cleverly extending a single-image depth estimator.

Trajectory Attention enhances video motion control by injecting trajectory information, improving precision and long-range consistency in video generation.

TeaCache: a training-free method boosts video diffusion model speed by up to 4.41x with minimal quality loss by cleverly caching intermediate outputs.

Open-Sora Plan introduces an open-source large video generation model capable of producing high-resolution videos with long durations, based on various user inputs.