Kwang Moo Yi
LightNews LightNews
kmyid.bsky.social
Kwang Moo Yi
@kmyid.bsky.social
57 followers 39 following 81 posts
Assistant Professor of Computer Science at the University of British Columbia. I also post my daily finds on arxiv.
Posts Replies Media Videos
kmyid.bsky.social
github.com/wustl-cig/pa...
GitHub - wustl-cig/parameter_interpolation
Contribute to wustl-cig/parameter_interpolation development by creating an account on GitHub.
github.com
November 27, 2025 at 8:21 PM
kmyid.bsky.social
arxiv.org/abs/2511.19811
Training-Free Generation of Diverse and High-Fidelity Images via Prompt Semantic Space Optimization
Image diversity remains a fundamental challenge for text-to-image diffusion models. Low-diversity models tend to generate repetitive outputs, increasing sampling redundancy and hindering both creative...
arxiv.org
November 27, 2025 at 3:03 AM
kmyid.bsky.social
ubc-vision.github.io/sonic/
And yes, it's our work ;)
November 26, 2025 at 3:06 AM
kmyid.bsky.social
github.com/valeoai/NAF
GitHub - valeoai/NAF: Official Implementation of NAF: Zero-Shot Feature Upsampling via Neighborhood Attention Filtering
Official Implementation of NAF: Zero-Shot Feature Upsampling via Neighborhood Attention Filtering - valeoai/NAF
github.com
November 25, 2025 at 9:27 PM
kmyid.bsky.social
arxiv.org/abs/2511.17454
Illustrator's Depth: Monocular Layer Index Prediction for Image Decomposition
We introduce Illustrator's Depth, a novel definition of depth that addresses a key challenge in digital content creation: decomposing flat images into editable, ordered layers. Inspired by an artist's...
arxiv.org
November 24, 2025 at 8:37 PM
kmyid.bsky.social
ai.meta.com/sam3d/
SAM 3D
SAM 3D can bring any 2D image to life, accurately reconstructing objects and humans, including their shape and pose.
ai.meta.com
November 21, 2025 at 8:35 PM
kmyid.bsky.social
github.com/Parskatt/rom...
GitHub - Parskatt/RoMaV2
Contribute to Parskatt/RoMaV2 development by creating an account on GitHub.
github.com
November 20, 2025 at 11:06 PM
kmyid.bsky.social
co-me-tokens.github.io
Co-Me: Confidence-Guided Token Merging for Visual Geometry Transformer
co-me-tokens.github.io
November 19, 2025 at 8:40 PM
kmyid.bsky.social
arxiv.org/abs/2511.11270
Φeat: Physically-Grounded Feature Representation
Foundation models have emerged as effective backbones for many vision tasks. However, current self-supervised features entangle high-level semantics with low-level physical factors, such as geometry a...
arxiv.org
November 18, 2025 at 8:19 PM
kmyid.bsky.social
github.com/FouierL/EquS
GitHub - FouierL/EquS: Official Code of the paper “Equivariant Sampling for Improving Diffusion Model-based Image Restoration“
Official Code of the paper “Equivariant Sampling for Improving Diffusion Model-based Image Restoration“ - FouierL/EquS
github.com
November 17, 2025 at 8:19 PM
kmyid.bsky.social
depth-anything-3.github.io
Depth Anything 3: Recovering the Visual Space from Any Views
depth-anything-3.github.io
November 14, 2025 at 9:25 PM
kmyid.bsky.social
time-to-move.github.io
Time-to-Move: Training-Free Motion Controlled Video Generation via Dual-Clock Denoising
time-to-move.github.io
November 13, 2025 at 7:54 PM
kmyid.bsky.social
arxiv.org/abs/2511.07743
UltraGS: Gaussian Splatting for Ultrasound Novel View Synthesis
Ultrasound imaging is a cornerstone of non-invasive clinical diagnostics, yet its limited field of view complicates novel view synthesis. We propose \textbf{UltraGS}, a Gaussian Splatting framework op...
arxiv.org
November 12, 2025 at 9:56 PM