AI Paper Atlas

Research map / Video / Video segmentation and tracking

LV-MAE: Learning Long Video Representations through Masked-Embedding Autoencoders

Ilan Naiman, Emanuel Ben-Baruch, Oron Anschel, Alon Shoshan, Igor Kviatkovsky, Manoj Aggarwal, Gerard Medioni

ICCV 2025 · Video › Video segmentation and tracking · approach: Self-supervised video representation learning · 4 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
add LV-MAE masked-embedding pre-training over raw embeddingsLanguageBind embeddings · LVU avgtop-1 accuracy: 38.9 -> 63.4Improved
add LV-MAE pre-training over raw embeddingsLanguageBind embeddings · LVU Writertop-1 accuracy: 5.95 -> 64.28Improved
add LV-MAE pre-training over raw embeddingsInternVideo2 embeddings · LVU avgtop-1 accuracy: 32.9 -> 59.6Improved
add MAE pre-training instead of end-to-end Mamba on clip embeddings (ablation)LanguageBind embeddings · LVU avgtop-1 accuracy: 56.8 -> 63.4Improved
add MAE pre-training instead of end-to-end Transformer on clip embeddings (ablation)LanguageBind embeddings · LVU avgtop-1 accuracy: 33.7 -> 63.4Improved
use LanguageBind clip embeddings instead of frame-MAE features (ablation)LV-MAE · LVU avgtop-1 accuracy: 54.3 -> 63.4Improved
use attentive probing instead of linear probing (ablation)LV-MAE LanguageBind · LVU avgtop-1 accuracy: 59.02 -> 63.41Improved
use semantic masking instead of random masking (ablation)LV-MAE LanguageBind (attentive) · LVU avgtop-1 accuracy: 62.54 -> 63.41Improved
use semantic masking instead of random masking (ablation)LV-MAE InternVideo2 (linear probing) · LVU avgtop-1 accuracy: 53.63 -> 52.60Hurt
increase encoder depth from 16 to 32 blocks (ablation)LV-MAE · LVU avgtop-1 accuracy: 59.73 -> 63.41Improved
increase encoder depth from 24 to 32 blocks (ablation)LV-MAE · COINtop-1 accuracy: 93.26 -> 91.8Hurt
use 5 s segments instead of 10 s segments (ablation)LV-MAE · LVU avgtop-1 accuracy: 61.43 -> 63.40Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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