AI Paper Atlas

Research map / Video / Video segmentation and tracking

NIRVANA: Neural Implicit Representations of Videos With Adaptive Networks and Autoregressive Patch-Wise Modeling

Shishira R. Maiya, Sharath Girish, Max Ehrlich, Hanyu Wang, Kwot Sin Lee, Patrick Poirson, Pengxiang Wu, Chen Wang, Abhinav Shrivastava

CVPR 2023 · Video › Video segmentation and tracking · approach: Neural video representations · 45 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 patch-wise autoregressive modeling instead of pixel-wise SIRENVideo INR compression · UVG-HDPSNR: 27.20 -> 34.71Improved
Use patch-wise autoregressive modeling instead of frame-wise NeRVVideo INR compression · UVG-HDencoding time (hours): ∼80 -> 6.71Improved
Use patch-wise autoregressive modeling instead of frame-wise NeRVVideo INR compression · UVG-HDPSNR: 37.36 -> 37.70Improved
Use NIRVANA instead of NeRVVideo INR compression · UVG-4Kencoding time (hours): ∼134 -> 20.89Improved
Use NIRVANA instead of NeRVVideo INR compression · UVG-4KPSNR: 35.24 -> 35.18Hurt
Use NIRVANA rather than a fixed-size NeRV model as video duration growsVideo INR compression · YouTube-8M Mario Kart, 2000/3000/4000 framesPSNR: 33.38 -> 31.6 -> 30.53 (NeRV); NIRVANA: 35.43, 35.21, 35.15Improved
Increase entropy regularization coefficient (ablation)NIRVANA · Jockey, UVG-HDBPPImproved
Increase entropy regularization coefficient (ablation)NIRVANA · Jockey, UVG-HDPSNRHurt
Increase patch size (ablation)NIRVANA · Jockey, UVG-HDPSNRImproved
Increase patch size (ablation)NIRVANA · Jockey, UVG-HDBPPHurt
Increase frame group size (ablation)NIRVANA · Jockey, UVG-HDBPPImproved
Increase frame group size (ablation)NIRVANA · Jockey, UVG-HDPSNRHurt

Later papers that compared with this method

1 later accepted paper reports NIRVANA as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 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