Computer Vision

BIR-D tames generative diffusion models for universal blind image restoration, dynamically updating parameters to handle various complex degradations without assuming degradation model types.

DoSSR: A novel SR model boosts efficiency by 5-7x, achieving state-of-the-art performance with only 5 sampling steps by cleverly integrating a domain shift equation into pretrained diffusion models.

SyncVIS: A new framework for video instance segmentation achieves state-of-the-art results by synchronously modeling video and frame-level information, overcoming limitations of asynchronous approache…

SyncTweedies: a zero-shot diffusion synchronization framework generates diverse visual content (images, panoramas, 3D textures) by synchronizing multiple diffusion processes without fine-tuning, demon…

Boosting diffusion model features: This paper introduces GATE, a novel method to suppress ‘content shift’ in diffusion features, improving their quality via off-the-shelf generation techniques.

SuperVLAD: A new visual place recognition method boasts superior robustness and compactness, outperforming state-of-the-art techniques by significantly reducing parameters and dimensions.

Real-time rendering of objects with subsurface scattering effects is now possible with SSS-GS, a novel method combining explicit surface geometry and implicit subsurface scattering for high-quality no…

Structured Unrestricted-Rank Matrices (SURMs) revolutionize parameter-efficient fine-tuning by offering greater flexibility and accuracy than existing methods like LoRA, achieving significant gains in…

StreamFlow accelerates video optical flow estimation by 44% via a streamlined in-batch multi-frame pipeline and innovative spatiotemporal modeling, achieving state-of-the-art results.

STONE: A novel submodular optimization framework drastically cuts 3D object detection training costs by cleverly selecting the most informative LiDAR point cloud data for labeling, achieving state-of-…

StepbaQ enhances quantized diffusion models by correcting accumulated quantization errors via a novel sampling step correction mechanism, significantly improving model accuracy without modifying exist…

START, a novel SSM-based architecture with saliency-driven token-aware transformation, achieves state-of-the-art domain generalization performance with efficient linear complexity.

Stable-Pose: Precise human pose guidance for text-to-image synthesis.

DiGIT stabilizes image autoregressive models’ latent space using a novel discrete tokenizer from self-supervised learning, achieving state-of-the-art image generation.

D³GM, a novel score-based diffusion model, enhances stability & generalizability in solving inverse problems by leveraging measure-preserving dynamics, enabling robust image reconstruction across dive…

SSDiff: A novel spatial-spectral integrated diffusion model for superior remote sensing pansharpening.

SSA-Seg improves semantic segmentation by adapting pixel-level classifiers to the test image’s semantic and spatial features, achieving state-of-the-art performance with minimal extra computational co…

SplitNeRF: One-hour training on a single GPU yields state-of-the-art scene geometry, lighting, and material property estimation!

Researchers introduce Video Gaussian Representation (VGR) for versatile video processing, embedding videos into explicit 3D Gaussians for intuitive motion and appearance modeling.

Spiking Experts Mixture Mechanism (SEMM) boosts Spiking Transformers by integrating Mixture-of-Experts for efficient, sparse conditional computation, achieving significant performance improvements on …