AI Paper Atlas

Research map / 3D vision / 3D generation

Rays as Pixels: Learning A Joint Distribution of Video and Camera Trajectories

Wonbong Jang, Shikun Liu, Soubhik Sanyal, Juan Camilo Perez, Kam Woh Ng, Sanskar Agrawal, Juan-Manuel Perez-Rua, Yiannis Douratsos, Tao Xiang

ICML 2026 · 3D vision › 3D generation · approach: Diffusion-based novel view synthesis

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
represent cameras as raxels (VAE-compatible RGB maps) instead of Plucker embeddings (ablation)Rays as Pixels · DL3DV-140 cycle self-consistencyFVD: 333.56 -> 68.17Improved
add decoupled self-cross attention (DSCA) (ablation)Rays as Pixels · DL3DV-140 cycle self-consistencyFVD: 77.08 -> 68.17Improved
add cosine similarity loss (ablation)Rays as Pixels · DL3DV-140 cycle self-consistencytranslation error: 0.094 -> 0.018Improved
add cosine similarity loss (ablation)Rays as Pixels · DL3DV-140 cycle self-consistencyFID: 9.48 -> 7.33Improved
increase diffusion steps for pose estimation from 2 to 20 (ablation)Rays as Pixels · DL3DV-140 pose estimationmRRA@30: 88.37 -> 85.90Hurt
use 1 diffusion step instead of 2 for pose estimation (ablation)Rays as Pixels · DL3DV-140 pose estimationmRRA@30: 88.37 -> 82.78Hurt
compare against VGGT pointmap-based pose estimatorRays as Pixels · Re10K pose estimationmRRA@30: 95.91 vs 98.07Hurt
use joint video-camera model for camera-controlled generationRays as Pixels · DL3DV-140 generationFID: 17.74 (Wonderland) -> 9.73Improved
use joint model vs Kaleido on trajectory adherenceRays as Pixels · DL3DV-140 generationrotation error: 0.011 (Kaleido) vs 0.098Hurt

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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