AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

GaussianVision: Vision-Language Alignment from Compressed Image Representations using 2D Gaussian Splatting

Yasmine Omri, Connor Ding, Tsachy Weissman, Thierry Tambe

CVPR 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Recaptioning and text-supervised pretraining · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use structured grid initialization instead of random for 2DGS fitting (4900 Gaussians) (ablation)2DGS fitting · 3000 iterationsPSNR: 28.24 -> 35.25Improved
use structured initialization instead of random (400 Gaussians) (ablation)2DGS fitting · 3000 iterationsPSNR: 17.77 -> 22.04Improved
batch-aware CUDA kernels2DGS fitting · batch 4096, 400 Gaussiansfitting speedup: 90.3XImproved
reduce Gaussian budget from 3136 to 400 splatsGS CLIP encoder · CLIP benchmark 38 datasetsrelative zero-shot accuracy: 0.98 -> 0.91Hurt
halve tokens from 196 to 98 for GS encoders (ablation)GS CLIP encoder · CLIP benchmark 38 datasetsrelative accuracy: 0.98 -> 0.96 (3136 splats)No clear effect
halve tokens from 196 to 98 for RGB by averaging patch pairsRGB ViT-B/16 · CLIP benchmark 38 datasetsrelative accuracy: 1.00 -> 0.87Hurt
replace RGB encoder with GS-3136 in LLaVA VLMLLaVA Vicuna-7B · six VQA benchmarksVQA score: MM-Vet 16.60 -> 19.40; POPE 77.83 -> 77.80Mixed
replace RGB encoder with GS-400 in LLaVA VLMLLaVA Vicuna-7B · six VQA benchmarksVQA score: SQA-IMG 63.81 -> 64.80; GQA 51.42 -> 50.21Mixed
use prune with larger initial Gaussian budget (ablation)2DGS pruning · Mini-ImageNet 224x224PSNR after pruningImproved
train CLIP from raw GS inputs directly instead of two-stage distillation (ablation)GS CLIP encoder · CLIP trainingconvergenceHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy