Computer Vision

Researchers significantly boosted the efficiency and quality of rectified flow, a method for generating samples from diffusion models, by introducing novel training techniques that surpass state-of-th…

Deep Fourier Shifting boosts small image restoration networks by using an information-lossless Fourier cycling shift operator, improving performance across various low-level tasks while reducing compu…

Boosting 3D point cloud recognition robustness, Frequency Adversarial Training (FAT) leverages frequency-domain adversarial examples to improve model resilience against corruptions, achieving state-of…

ImOV3D: Revolutionizing open-vocabulary 3D object detection by learning from 2D images alone!

Immiscible Diffusion boosts diffusion model training efficiency up to 3x by cleverly assigning noise to images, preventing the mixing of data in noise space and thus improving optimization.

IMAGPose: A unified framework generating high-fidelity person images from single or multiple source images & poses, addressing existing methods’ limitations.

Leveraging image understanding models for image tokenizer training dramatically boosts image generation quality, achieving state-of-the-art results.

aSeqDIP: A new unsupervised image reconstruction method using sequential deep image priors, achieving competitive performance with fewer data needs and faster runtimes.

ICDiff, a novel Image Copy Detection system, tackles the unique challenge of identifying replicated content in diffusion model outputs, introducing a specialized dataset and deep embedding method for …

IllumiNeRF: Relightable 3D reconstruction without inverse rendering using image diffusion and NeRF.

MuDI: a novel framework for multi-subject image personalization, effectively decoupling identities to prevent mixing using segmented subjects and a new evaluation metric.

Researchers solve the conditional image leakage problem in image-to-video diffusion models by proposing a new inference strategy and a time-dependent noise distribution for training. This yields video…

ID-to-3D: Generate expressive, identity-consistent 3D human heads from just a few in-the-wild images using score distillation sampling and 2D diffusion models.

Hyper-SD boosts diffusion model speed by using trajectory segmented consistency distillation and human feedback, achieving state-of-the-art performance.

HydraViT: Stacking attention heads creates a scalable Vision Transformer, adapting to diverse hardware by dynamically selecting subnetworks during inference, improving accuracy and efficiency.

Hybrid Mamba Network (HMNet) boosts few-shot segmentation accuracy by efficiently fusing support and query features using a novel hybrid Mamba architecture, significantly outperforming current state-o…

HumanSplat: single image-based 3D human reconstruction using Gaussian Splatting with structural priors, achieving state-of-the-art quality and speed.

DIFFUSIONHOI: A novel HOI detector using text-to-image diffusion models to improve compositional reasoning and handling of novel concepts, achieving state-of-the-art performance.

Human-3Diffusion generates realistic 3D avatars from single RGB images using coupled 2D multi-view and 3D consistent diffusion models, achieving high-fidelity geometry and texture.

Inline Prior Guided Score Matching (IPSM) improves sparse-view 3D reconstruction by leveraging visual inline priors from pose relationships to rectify rendered image distribution and effectively guide…