Research map / Video / Video generation
EgoX: Egocentric Video Generation from a Single Exocentric Video
Taewoong Kang, Kinam Kim, Dohyeon Kim, Minho Park, Junha Hyung, Jaegul Choo
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add geometry-guided self-attention (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 14.77 -> 16.05 | Improved |
| add geometry-guided self-attention (ablation) | EgoX · Ego-Exo4D (seen scenes) | FVD: 254.08 -> 184.47 | Improved |
| add egocentric prior (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 13.67 -> 16.05 | Improved |
| add egocentric prior (ablation) | EgoX · Ego-Exo4D (seen scenes) | Dynamic Degree: 0.802 -> 0.974 | Improved |
| add clean latent conditioning (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 15.07 -> 16.05 | Improved |
| add clean latent conditioning (ablation) | EgoX · Ego-Exo4D (seen scenes) | FVD: 343.33 -> 184.47 | Improved |
| use channel concat for ego prior and width concat for exo (full) instead of prior width, exo channel (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 13.83 -> 16.05 | Improved |
| use width+channel unified conditioning instead of both width-wise (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 14.85 -> 16.05 | Improved |
| train with GGA rather than applying it only at inference (ablation) | EgoX · Ego-Exo4D (seen scenes) | PSNR: 15.23 -> 16.05 | Improved |
| add geometry-guided self-attention (ablation) | EgoX · Ego-Exo4D (unseen scenes) | PSNR: 13.27 -> 14.38 | Improved |
| add egocentric prior (ablation) | EgoX · Ego-Exo4D (unseen scenes) | PSNR: 13.01 -> 14.38 | Improved |
| add clean latent conditioning (ablation) | EgoX · Ego-Exo4D (unseen scenes) | FVD: 695.01 -> 440.64 | Improved |
Later papers that compared with this method
1 later accepted paper reports EgoX as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Grounded-Exo2Ego: Structured Semantic Grounding for Robust Exocentric-to-Egocentric Video Generation (NeurIPS 2026 · better on 8 of 8 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
- Exocentric-to-Egocentric Video Generation (NeurIPS 2024 · similarity 0.93)
- Learning Egocentric Cues from Exocentric Video using Privileged Egocentric Supervision (ECCV 2026 · similarity 0.91)
- EgoVid-5M: A Large-Scale Video-Action Dataset for Egocentric Videos Generation (NeurIPS 2025 · similarity 0.91)
- Grounded-Exo2Ego: Structured Semantic Grounding for Robust Exocentric-to-Egocentric Video Generation (NeurIPS 2026 · similarity 0.90)
- EgoGen: An Egocentric Synthetic Data Generator (CVPR 2024 · similarity 0.90)
- Retrieval-Augmented Egocentric Video Captioning (CVPR 2024 · similarity 0.89)
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.
